December 4, 2024 • 57 mins
What if you could simplify your entire networking infrastructure with one holistic solution? Join us in this episode as we sit down with Anil, the visionary CEO of Meter, to uncover how his company is revolutionizing the networking industry. With a fresh perspective on hardware and software integration, Anil shares his journey from a network engineer to a leader at the forefront of innovation, offering products that give engineers complete control and visibility over their networks. Discover Meter's unique offerings, like the Connect marketplace for ISPs and intuitive natural language commands that are set to transform user experience.
We explore the strategic advantages of being a startup in a field dominated by legacy players. Anil reveals how Meter sidesteps the technical debt faced by older companies, providing superior products and aligning incentives with customer needs. By taking on capital risks and eliminating hardware costs, Meter distinguishes itself from traditional vendors who focus on hardware sales margins. Listen as Anil discusses the long-term contracts that ensure a return on investment and the unwavering commitment to exceptional service that sets Meter apart.
Anil explains the significance of trust when integrating new vendors into existing systems and highlights the innovative ISP marketplace feature that eases the management of network services. Addressing common concerns about automation and cloud technologies, Anil reassures listeners that these advancements are meant to enhance, not replace, jobs.
Tune in to learn how Meter's solutions, from streamlined network service management to cost-saving recommendations, are redefining the landscape of network engineering and education.
For more info on Meter's solutions:
www.meter.com
Connect: www.meter.com/connect
Command: www.meter.com/command
Find everything AONE right here: https://linktr.ee/artofneteng
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
This is the Art of.
Speaker 2 (00:01):
Network Engineering podcast.
In this podcast, we'll exploretools, technologies and talented
people.
We aim to bring you informationthat will expand your skill
sets and toolbox and share thestories of fellow network
engineers.
Speaker 1 (00:23):
Welcome to the Art of Network Engineering.
My name is Andy Laptev andtonight we have a very special
episode.
This is a sponsored episodewith the good folks at Meter.
Meter is a networking as aservice company.
They are making really coolstrides in the NAS business and
(00:44):
we're going to do a few thingslike define what that is and see
what kind of problems they'resolving and how they're doing it
.
Anil is a really cool guy.
He's a career network engineer.
He's one of us people, so Ithink that that's really
important.
If you're going to come on theshow, you better be a network
engineer talking about yourproduct.
So Anil has the credentials.
(01:04):
We're also joined by ChrisMiles in the future in Australia
.
How are you doing, chris?
Speaker 2 (01:08):
Doing well.
Brother, Thanks for having me.
Speaker 1 (01:09):
Absolutely.
Thanks for coming on.
So, anil, let's tell folksabout yourself.
Who are you, what do you do?
How did you get to be a networkengineer to this CEO of a NAS
company?
Speaker 3 (01:25):
Andy, chris, thanks for having me and thanks
everybody for tuning in.
I studied networking in collegeand I've always felt networking
is one of the most importantthings in the world.
And maybe a career bet mybrother and I who I started
meter with made about decade anda half, two decades is that we
all will use internet more thanwe currently do, and we thought
(01:48):
that was a pretty safe bet tomake at that time and we
continue to believe that's avery safe bet to make today.
And everything that we do thispodcast or doing anything it's
all just packets somewherefloating around through wires
between radios, and we wanted tohave an opportunity to build
(02:11):
really great hardware, reallygreat software for networking,
and so meter at the core is justa networking.
We believe in a couple ofthings that I think the industry
lost a little bit of track of,especially as users and
customers.
We felt like everybody thoughthardware was commoditized,
nobody was really payingattention to software and nobody
(02:33):
was integrating productstogether.
So what we set out to do is,instead of building a point
solution, instead of sayingwe'll just build access points,
somebody else will figure outthe rest, we'll just build a
security product, somebody elsewill figure out the rest.
We'll just build a securityproduct.
Somebody else will figure outthe rest.
We said why don't we take thetime and build really good
products for the entire stack?
And so today, meter buildseverything from routing,
(02:56):
switching, wireless PDUs, ispmarketplace that's built in
SD-WAN, vpn, all of it wrappedtogether along with applications
and software to give networkengineers really good control
and really good visibility tothe level that I would want it
(03:18):
as a network engineer.
You mentioned networking as aservice.
We don't really think ofourselves as a networking as a
service company.
Oh, I apologize.
No, no, you don't need toapologize.
Networking as a service issomething we do, but at the core
, what we really believe is onnetworking itself across
(03:38):
hardware, software, operations,and the four big product lines
that we have that do this is.
One is our ISP product that'scalled Connect.
Think of it as Kayak or Expediafor ISPs.
How it started is that we foundit really intriguing that you
couldn't buy the internet on theinternet, so that's how it
(04:01):
started.
Then we have our wired productsthat are routing, switching.
Then we have our Wi-Fi productand about a month ago we
announced our cellular productand what we're hoping is in any
space if there's a packet moving, it goes through meter,
hardware and software and reallyfrom designing the hardware
(04:32):
operating systems, firmware,distributed systems, apis,
applications we just think itcan be way better If there's a
company that really cares aboutthe hardware, cares about the
software and integrating it alltogether.
We got started about a decadeago.
We've been relatively quietbecause we wanted to start
talking about meter when we knewwe had the best products for
networking.
So that's why you probablyheard more about us in the last
(04:52):
couple of years, in particularin the last year, because we
feel like we've crossed thatthreshold.
So that's who we are, that'swhat we do.
Speaker 1 (04:59):
That's a huge like you've integrated the entire
stack.
That is a big.
That is a big task.
You are you are you are verybrave and I don't know why, as
you were talking, I kind ofthought of just you know, I'm
trying to think of otherintegrated solutions and, like
Apple came to mind, right, theyhave the hardware and the
software and they're able toprovide a more integrated, more
(05:20):
secure, better experience than abunch of disparate parts and it
kind of sounds like that'swhere you're going is it's?
it's your hardware, it's yourNOS, it's your people, it's it
sounds like a good.
Give a much, much cleaner, um,I guess, experience for for your
customers.
So so who?
Who are your customers?
(05:41):
Right, like I'm a networkengineer.
So is this something?
I mean I guess you're marketingto businesses, correct?
People who need networking,infrastructure and things
installed.
Right, how does someone like me?
Right, it's all about me.
I want to know how this isimpacting our network engineers.
So we're going to get into theproduct.
I like I know that you have.
(06:04):
What's the new?
You have a really cool naturallanguage feature.
Yeah, command, which I'm reallyinto, because the more I see
that in networking platforms, Iget really excited because, you
know, as fun as it is learningarcane syntax in a bunch of
different CLIs, right, which wasthe whole job for a really long
(06:24):
time I love the fact that I canjust speak to the
infrastructure in naturallanguage and get insights.
I think that that's supervaluable.
You know, there's a lot ofAI-ification of everything,
right, and we've almost becomekind of immune to it, like oh,
great AI.
But when I saw the demo up onyour site, I'm like, oh, this is
(06:51):
like, this is real, this isuseful, I can ask a question and
get really valuable insights.
It looked like I could build, Ithink, like my own dashboards
too, based on the data that.
So I think I know this is amainly audio podcast and I think
that we should show some stuffat some point, you know, during
this.
I know it's still early.
I'd like to see, but I guessback to my original question.
So let's say, I'm a businessand I'm starting up and I need
internet, right, I needinfrastructure.
(07:11):
So I call you guys and what doyou do for me?
I think you do everything.
Speaker 3 (07:16):
Yes, so our customers are folks like you mentioned
that are starting up newbusinesses, all the way up to
some of the largest companies inthe world, so large public
companies.
A lot of network engineers arelistening when we all congregate
on the internet.
If you imagine the form of theinternet that runs on meter.
(07:39):
The largest hedge fund in theworld, bridgewater runs on meter
.
Whole gamut of companies.
Actually, chris works at acompany that runs on meter,
another networking company thatI just found out before we hit
record and so our customers andwho we're building for is
entirely for network engineersnetwork engineers and the
(08:08):
premise of our business when westarted it was precisely that
nobody was building products fornetwork engineers anymore and
that felt very obtuse.
We felt that as users ourselves.
It felt like random products,random features.
You know forgive me, andy, butcrazy marketing stuff that
didn't make any sense and we'rejust like where are the real
(08:29):
products?
Who are the people that reallylove this?
So our products are built fornetwork engineers.
Our customers range frombusinesses that are just getting
started up to the largestbusinesses in the world that
have the most security,compliance requirements and
really large technology andnetworking teams, and when
(08:50):
somebody signs up.
Our work starts with firstmaking sure the right ISP gets
to this business, because nomatter how great of a network
you have inside of a space, ifyou don't have great when coming
in and out, kind of pointless.
So that's where our Connectproduct comes in, which is
exactly that of Kayak or Expedia.
You just literally type in anaddress, we'll show all the ISPs
(09:14):
you can purchase and you canpurchase it instead of waiting
two weeks of going back andforth to figure out who's
available and what price.
And we actually made thatentirely a free product because
we wanted to also contribute tothe community.
So anybody can go to metercomslash connect and just use it.
Then we work with our partnersso some of the largest channel
(09:37):
partners to some of the smallestones or us that understand a
space, and we design a network.
So both we can work with thechannel or work directly with
the customer to design a network.
Then it's our rack that'sinstalled entirely, that has
PDUs, routing, switching andeverything you can imagine there
(09:59):
, from high availability to hostmonitoring and DNS security and
stateful firewall and all thosethings that are built in that
you would expect that are deeplyintegrated together, along with
indoor and outdoor wireless.
And then now, as I mentioned, acellular product that, instead
of you having to go spend a lotof money on DAS and other
(10:21):
systems, cellular deployed justlike Wi-Fi, we deploy that.
Then we give what we hope isbeautiful software with our
dashboard and things likecommand that you mentioned, that
give full control to ourcustomers, who are, again,
primarily the networking teamsat the businesses that we work
(10:42):
with.
Speaker 1 (10:42):
So quick question is this how do I interact with the
infrastructure?
Is this a GUI?
I'm guessing I have an API onthe backend.
Speaker 3 (10:50):
Yeah, just primarily the dashboard and primarily
command.
Command.
Like you mentioned, we launchedit about two months ago and
it's been an incredible success,not just in networking but in
technology as a whole.
Many people thought it was oneof the best applications and
models that people had ever seen.
So real software directlyavailable to customers.
Speaker 1 (11:14):
So we don't have to learn a new CLI, right?
Yeah, yeah.
Speaker 3 (11:17):
No, yeah, that's where command comes in.
It's beautiful, and one of thethings about software that's
challenging is either you'redesigning it for the novice or
you're designing for the expert.
When you start a new softwareproject, you're kind of told
pick one.
And I think what's beeninteresting about command that
(11:37):
we built entirely in-house it'snot APIs, we're calling
elsewhere and it doesn't go tosomething like OpenAI or it's
all meter is command is flexibleand kind of goes around,
depending on the user, and thepremise of command is that
software should be soft, and so,along with the dashboard that's
(11:59):
fully fleshed out, command cando things like give information
for anything you're asking, takeaction.
It can be really complicatedaction.
Like you know, create me a newSSID with this VLAN and make
sure it auto rotates every dayat 2pm and put it on all the
access points.
You can actually say that andthen we'll go do it, and then
(12:20):
command also just generatessoftware on the fly, just the
way you want it.
We want every one of ournetwork engineering and IT teams
to have a software engineerjust for them, but it's a full
software platform integrateddeeply into the hardware.
Speaker 1 (12:37):
So one of the reasons I love live streaming is we
have a live audience and I mighthave just found you a customer
because we have someone in thechat saying wait, did I hear
right A DAS replacement?
I might be interested.
We're looking for a newsolution.
I am not familiar with DAS.
I'm Googling it on the side sowithout me giving a terrible
(12:58):
explanation?
Do you want to hit a high level, like what is DAS, and why
would this gentleman be smart ingetting it to you, sure.
Speaker 3 (13:04):
So DAS stands for Distributed Antenna System.
So when you go inside of abuilding today, when you open up
your phone, the goal is bothfor the carriers and IT teams is
that you should have five barson your cell phone.
But that's actually really hard, challenging and expensive to
do On the carrier side.
Carriers have to go installantennas, rent out rooftops and
(13:27):
spend a lot of money, and thencompanies install these systems
called DAS systems.
If you pull up an image, go toGoogle Images, you'll see there
are these antiquated systemsthat look like servers in Matrix
that go in basements and thenthey have antennas that are
distributed throughout abuilding and so when you pick up
(13:49):
your phone, your phone is goingfrom you know signal from your
phone to one of these DASantennas to one of these outdoor
antennas that are popped up,and then it just meshes all the
way through next stop, next stop, next stop to get to whichever
carrier that you are.
What meter cellular enables youto do is a couple of really
(14:10):
interesting things.
Just like there's a Wi-Fi AP,there's now just a cellular AP
and it's connected to the fullnetworking stack that we built.
This is the advantage ofbuilding integrated products.
And now, when you open up yourphone, instead of it going from
your phone to an indoor antenna,then hopping all these external
antennas, it actually just goesfrom your phone to the cellular
(14:33):
AP and directly throughinternet to whichever carrier
you have.
So you, as a customer, can getgreat cellular that's deployed
in 75% less time, with somethinglike 80% less of the cost, and
cellular deployed as easy asWi-Fi that's again integrated
(14:54):
into the entire networking stackthat a company already needs.
Speaker 1 (14:58):
All right, so hopefully our listener Nully
heard that.
They should reach out to thesefolks.
It sounds like they have areally good solution.
Speaker 3 (15:07):
Yeah, shoot me an email, Anilatmetercom.
I'm super simple.
Speaker 1 (15:11):
I never thought about , I guess, cellular in large
buildings and how difficult thatwould be.
I mean, I know how hard it isto get Wi-Fi in buildings.
But yeah, the cell towers areoutside.
You know, you don't havetypically access points inside.
Speaker 2 (15:22):
So if you see those little knobs on the ceiling that
aren't APs.
That's typically what those are.
Speaker 1 (15:27):
They're little antenna repeaters basically.
Speaker 2 (15:31):
I do have a question Now.
You've kind of already touchedon this and I think I know where
you're going, but I want togive you the opportunity to
expand on it.
So you know, if I think aboutthe landscape of networking
vendors today, obviously there'sa lot of them, or there's a lot
out there, there's a lot thathave been around for a very long
time and you know they probablyhave claims as well that they
(15:53):
do everything in the stack, kindof similar to what we're
talking about today.
But my question is you know, ifI think about your all stance,
obviously you're relatively newto the game.
You've already made quite a bitof headway.
Your all stance, obviouslyyou've, you've, you're
relatively new to the game.
You've, you've, um already madequite a bit of headway.
Do you feel that being astartup in this scenario kind of
helps you guys with umreleasing better products, just
(16:13):
because you don't have the yearsof technical debt that maybe
some of the other people do inthis space?
And can you maybe elaborate onon what your, what your uh input
is there?
Speaker 3 (16:23):
yeah, we've only been around 10 years where, like you
mentioned, there's others thathave been around 20, 30, 40
years.
So totally we've only beenaround 10.
We've not been around as longas some of the other folks.
I think it's less technicaldebt because of the years.
(16:45):
I actually think it's moreabout our business model and how
our incentives are aligned withour customers that enable us to
do things that others can't.
So a couple of things that I'llmention here what that means.
One is meter builds an entirelyintegrated stack.
We don't sell individual piecesof hardware, so we don't just
(17:06):
sell an AP.
We don't sell individual piecesof hardware, so we don't just
sell an AP.
We don't just sell a switch.
We sell networking.
What that enables us to do isalways know what's above us in a
stack and what's below us andhave the right APIs, right
protocols and right thetransactions between each piece
of hardware to be able to buildgreat networks.
It enables us to make differentchoices on hardware.
(17:28):
We are not expecting our APs toalso do some routing and some
switching.
Our APs only do wireless andthat leaves a lot of room on the
chipsets, because we all usethe same Qualcomm chipsets.
We use the latest, greatestQualcomm.
So does everybody else.
But it allows room to go domore and really focus on
(17:50):
wireless really focus onwireless.
And our business model also isthat we take on the capital risk
entirely instead of ourcustomers.
What that means is there's zerocost for the hardware,
including when we release newhardware.
So every four or five yearsthere's a new Wi-Fi spec for my
triple E, or when there's newswitching algorithms or routing
(18:12):
or new hardware, we make newhardware.
Or when there's new switchingalgorithms or routing or new
hardware, we make new hardware.
All of our customers getupgraded to that new hardware as
well.
But what that means with ourbusiness model that we take on
the capital risk and hardwarecosts are zero is that we only
get paid if we provide a goodproduct and a good service to
our customers.
That means our incentives withour customers are very much
(18:32):
aligned.
We are not looking to make abox and sell that for 25% margin
and hope to sell you a new boxevery 18 months.
25 is low, I was being generous.
But our business model isentirely aligned with our
customers.
And if you take a step back andthink about it, networking is
(18:55):
the last part of the technologystack that, until meter came
along, wasn't aligned withcustomers.
Everything from compute tostorage, to CRMs, to everything,
was that if you provide me agreat product, that's when you
get paid.
But networking was that you payme upfront.
And great product that's whenyou get paid, but networking was
that you pay me upfront andthen you'll never hear from me
(19:17):
again until I want to sell youanother thing.
So that actually enables us todo much different things on how
we can integrate hardware,integrate software and build
products other people can't.
Because we know what's above usand below us in the stack,
because each part of our stackis built to do one thing that
(19:37):
it's good at and great at, andhow our business model is set up
so that our incentives arealigned with our customers.
Speaker 1 (19:46):
Yeah when we were originally talking.
We did an episode a while backon networking as a service and
somebody in the chat saidnetwork attached storage, no,
the other NAS.
But I wondered how?
Because hardware is soexpensive at least traditional
hardware that we're used to Ihad thought how do you get a
(20:07):
return?
Like for you to just hand metens of thousands of dollars of
hardware.
I'm assuming I'm signing acontract because you need to get
a return, I guess, on all thehardware you just handed me.
So is it safe to say that we'rein some type of contract so
that you can get your return?
Speaker 3 (20:24):
Yeah, not just tens of thousands of dollars.
With our largest customers thathave tons of locations, square
footage and a lot of footprint,millions and tens of millions of
dollars of hardware, and Idon't mean to make it sound
different than the traditionalmodel, because you're in the
same boat.
Speaker 1 (20:40):
If you spend $15 million with a big vendor, right
?
You're in it for five to sevenyears, and whether you have a
good experience or not.
So it's not a downside of this,but I'm just wondering.
I'm expecting some type oflong-term relationship, just so
you can it is long-term Three,five, seven years are generally
what we see from customers.
Speaker 3 (20:59):
But what's implicit in all of these contracts that
we do with our customers is thatwe're providing them great
products and great services, andthat's where it's different
than before, where you're notjust giving us the money and
then that's the end of it.
We have to earn the right tohave you as a customer, but it
is definitely something becausewe believe in building the right
(21:21):
products.
It's definitely a bet we'remaking that we will get to keep
the companies as customers, butalso those networking teams, no
matter where they go, ascustomers as well, because we
would like to obviously buildthe largest networking company
possible, and not only do wewant to keep the companies, but
also the people forever.
Speaker 1 (21:41):
How do we get you deployed?
Is this Greenfield Good?
Speaker 3 (21:45):
question.
About 40, 45% of what we do isBrownfield, and what we also
have is a buyback program, sothere's no sunk costs.
Let me break that down a littlebit Generally.
When an IT team has to make adecision, they have to work with
finance, they have to work withprocurement, they have to work
(22:05):
with vendor management offices.
It's not just a technicaldecision, it's a business
decision, and they have othercolleagues and departments that
they owe an explanation to whichis hey, we just spent $10
million on networking gear.
What's happening?
You want to use this new thing?
What happened to all that moneythat we just spent?
Or you just spent 100K.
What happened to it?
(22:26):
So we actually buy out all thelegacy hardware that companies
have to make the switching costseasier over the meter as well,
so that it's much more palatablefor procurement, finance and
other teams.
And again, we serve IT andnetworking teams and our job is
to make their lives easier.
Speaker 1 (22:47):
I'm guessing that you're also.
You're attractive to thefinance departments, right is I
know?
That a lot of the as a servicestuff is, you know, opex versus
capex, right?
So that's?
I mean that's this model right,I have a recurring bill we have
both models some cost, oh okaywe have both models.
Speaker 3 (23:04):
Um, you know, we have customers that are owned by pe
firms.
Opex hurts their eDA margins,so we have capital lease models
for those folks as well, and wehave pure OPEX models as well.
But what's inherent in both isthat we're taking the capital
risk on the hardware, not ourcustomers.
So our customers get to havethe benefit of the fact that
(23:29):
Meter is taking the risk on thehardware, that meter is taking
the risk on the hardware andmeter is on the hook to provide
really great products andservices to customers to earn
and continue to earn theirbusiness.
Speaker 1 (23:40):
You told me to ask hard questions, so I'm going to
Please Otherwise it's not funfor me either.
Well, no, but I think it's fair.
So I'm following you and so farthis is compelling.
I have experienced qualityissues right With different
vendors software, hardware.
So how's your quality, if youcan make a generalized statement
(24:02):
?
You know, I'm guessing you havesupport for RMAs, right.
Hardware is going to fail.
Nothing's perfect Because it'sintegrated.
I'm assuming it's a very tightintegrated experience and the
assumption I'm making is thatthe quality is high because I've
seen your customer list.
It's large.
If you weren't doing a good job, people would be yelling about
(24:23):
it on the internet, becausethat's what people do on the
internet is they yell ateverything they're mad about,
and I've never heard or seenanyone yell about how mad they
are at meters.
So can you talk about yourquality?
Just your NAS, your hardware,things that would be important
for folks.
Speaker 3 (24:36):
Yeah, I think.
Before jumping into specifics,which I will, I want to tie it
back to what I was mentioning onthe business model itself.
What's tied to quality isactually speed to market, and
one of the challenges we thinkthe legacy vendors have with
their model is I'm sure you'veseen this in marketing from all
(24:57):
these companies they want to andneed to say world's first Wi-Fi
7 AP.
They all want to and need toAbsolutely For a lot of
different reasons.
They need to be first to market.
They're like we're doing Wi-Fi7.
I'm going to put this randomchip in there, this random
protocol, just because I'mtrying to get Andy to buy one
(25:19):
more time, even though he justbought three years ago.
I need to convince him that heneeds this new box.
Speaker 1 (25:25):
It's that new iPhone.
Every two years I need the newone.
That's right.
Speaker 3 (25:29):
We don't have that affliction.
You will never see us beworld's first.
In fact, those are kind of thesilly statements you will never
hear me to say like world'sfirst Because it doesn't matter
for world's first, does itactually work?
Does it work for me?
So again, it ties back to ourbusiness model is we don't have
to and we don't need to go dothose silly games and play
(25:52):
finite games.
We like to play infinite gamesthat keep going forever.
Second is, like you're saying,because we have this business
model and this verticalintegration and we sell the
entire network, there's threeimportant things that affect
(26:14):
quality.
One is a unified operatingsystem across the entire
platforms is a unified operatingsystem across the entire
platforms and what that buys usis the ability to know that, oh,
ap version this and firmwareversion this for firewall
doesn't work or some feature.
Those issues don't occur.
Actually, when people thinkabout quality issues, they go to
code quality and testingquality issues.
Yes, those things happen, butin reality, the actual issues in
(26:38):
networking for quality comefrom integration issues.
Even with the same vendor,which is you might have this
version of the switch not workwith this version of AP, not
work with this version offirewall and this one feature
wasn't tested properly becauseof that, those kinds of things.
Because we have a single unifiedfirmware image and operating
(26:59):
system, we know what actuallyworks together and those are
releases that are cut that wayevery quarter with that in mind.
What it also buys us is asingle API across the entire
stack, and how that shows up isboth in quality but also in
usage.
When you want to make changes,you can make change in one part
(27:23):
of the stack and actually haveit traverse everywhere, really
simply, and enable us to buildgreat software for our customers
because of it.
And the third thing it buys usis also a unified stats pipeline
actually knowing how the entirenetwork is doing throughout the
pipeline of a packet, from ISP,routing, switching, wireless
(27:47):
through firewall, through DNSsecurity, through VPNs, through
SD-WAN.
All of it is connected with aunified stats pipeline.
That actually was enable us tobuild much higher quality
products because of the stancewe took on the entire network,
because of our business modeland how we're set up to build
hardware and software.
Speaker 1 (28:09):
You just made me think of troubleshooting.
So I have a problem.
People are calling in, theycan't reach an application.
Let's say I'm thinking thetraditional troubleshooting I
would do.
I do some captures, some PCAPs,right.
So how do I troubleshoot this?
Is it any different From whatyou just described?
It sounds like I have insightsthat's right, Just at my
(28:30):
fingertips.
Speaker 3 (28:30):
Yeah, so you can do all the traditional stuff from
understanding everything fromclient health metrics to packet
captures.
But it's a unified thinginstead of you having to go look
at every part of the stack.
It's a huge advantage beingable to get that.
That's a push from the systemrather than a pull from every
(28:51):
part from you as an engineer.
Second, you referenced commandearlier.
One of the things you can ask.
Command is just like Slack orTeams.
You can actually just hit atand get to a location or get to
a meter hardware or get to aclient device.
So you can just hit at and sayAndy's Windows let's say that's
(29:12):
your computer's name and saytell me the health of this
client and give me a five-bulletsummary.
And at the speed of a Googlesearch it will go and summarize
all the things which you can golook at yourself, because you
have logs available andconnection history and all those
things available, and just gethere's the five bullets.
(29:34):
Hey, his snr kind of sucks, oryou know, andy wasn't supposed
to really use 2.4, but he'sactually on 2.4, um, those kind
of things old school man 2.4forever.
Speaker 2 (29:45):
So it sounds like something andy would do,
probably.
How dare you five five?
Speaker 1 (29:50):
gigs, five gigs, forever.
I.
The more you talk, the morequestions I come up with.
So, please, you know we welland Chris jump in, jump in
whenever.
So, can I, can you do cloudpeering?
So, like somebody in the chatsaid, you know they're curious
if BGP supported it.
So that just made me think,okay, well, can I, can I peer to
(30:10):
my CSPs from your infra?
And then what does that looklike?
You know, A, can we connect?
And then B, are there anyinsights there that you can pull
in?
Like again, something's broken?
I'm going back to my productiondays.
Yeah, I railed, I cried against, you know, about cloud, because
when something that was hostedin the cloud broke, it was very
(30:33):
confusing and complicated tofigure out which cloud it's in.
How is that routing built?
What exactly broke?
It was just magic somewhereelse.
So I guess we'll take it one ata time.
Do you?
Can I peer with CSPs withpublic cloud with your?
Speaker 3 (30:47):
Yeah, so we can.
We can connect to a pops publiccloud, direct connect, all
those things.
But I want to also answer thesecond question you had, which
is do we have visibility andobservability and what's
happening there Once you doconnect?
We have it on the connectionsitself, but not beyond.
Those are things we will buildlater on, but we don't have that
(31:07):
today.
I want to be clear that I'm not.
Speaker 1 (31:09):
And I wouldn't expect it to.
That's not even like a fair ask, but I'm just thinking all the
things you connect to right.
Yes, so you connect to cloud,which is good, that's important.
I mean, you guys have thoughtall this through.
It's not like, oh well, we onlydo Wi-Fi in-house and you got
to figure the rest out like youguys.
Speaker 3 (31:23):
No, no, routing, switching, security ISPs.
Speaker 1 (31:30):
And I like that.
It's I guess it's one bill.
I don't care when the billcomes, I'm assuming monthly, not
that it matters, but where I'mgoing with this is licensing.
One of the things that hasalways really made me nuts is,
you know how do I say this?
Subscription licensing.
I don't want to say anybody's,you know patented license term
and then Make it look like I'mtalking trash on that particular
(31:52):
vendor because I'm not.
But the subscription licensingmodel as a person who came up
again it's.
It's tough being around andbeing an old school person,
because back in my day this ishow it was done and it was
better, you know, when I had aperpetual license and that thing
was good forever.
I really enjoyed that, becauseI've been hit with outages when
subscription licenses expiredand we weren't paying attention
to the logs and bad thingshappen.
Speaker 2 (32:14):
So but I'm guessing you're subscribed.
The phrase I heard a lot wasone throat to choke.
I don't know if that's reallywhere you want to go with that,
but I heard that a lot Totally.
Speaker 1 (32:29):
But I don't get the sense from Meter that you know
like.
Licensing isn't going to be aproblem because you're not
hitting me with this separatemonthly bill to keep everything
working and the features up.
It's all integrated, it's onebill everything you need.
Yeah, which I really like.
That's even just simplifyingthe operational model.
Just that alone.
I know it sounds it should betable stakes, like your
licensing model shouldn't bringinfrastructure to its knees and
(32:51):
people at vendors will say, ohno, it doesn't we send log
messages and warnings.
And people at vendors will say,oh no, it doesn't.
We send log messages andwarnings.
And well, okay, yeah, becausewe're all sitting around
watching logs for expiringsubscription licenses, like
right, we're understaffed, we'renot, we're doing other things,
so okay, so, licensing is easy.
It's integrated Code.
Upgrades is another thing thatcame up for me, did you?
Say you release quarterly F.
Speaker 3 (33:11):
Firmware is quarterly , for dashboards and other
things weekly, sometimes daily.
We have a really nice velocity,but in general it's more driven
by customer outcomes.
One of the things we reallycare about which I think also
the industry has kind of lost alittle bit is, as we started
(33:33):
doing more fancier and fancierfeatures.
I think the basic networkingstuff wasn't done to perfection.
So that's the other thing wetry to do is we try to do the
basics really right, like mybrother who I started meter with
.
If you guys have them on andsay like, hey, tell me about STP
(33:54):
and how it should beimplemented, not used, be
implemented, not used,implemented That'll be a fun
conversation.
So we try to do the basics right.
The software is derived fromthe outcomes from customers and
really just going everythingderiving from those interactions
with customers and howcustomers get to use our
(34:16):
hardware and our software.
Because at the end of the day,I think IT and networking teams
have one of the toughest jobs inall of any industries actually,
because they're only thought ofwhen something doesn't work.
When the network is working,nobody's calling a networking
guy or gal saying, hey, thenetwork is fantastic, good job.
(34:38):
Never, it's never happened.
It's only when it doesn't work.
Speaker 2 (34:45):
Proving innocence rather than that's right
Function.
Speaker 3 (34:48):
That's right.
That's right.
So all the software that webuild code upgrades, how we do
firmware, how we do testing, howwe do firmware, how we do
testing, how we do dogfooding,how we do canary upgrades.
All of that is tied back totrying to understand the lives
of our customers, whichpredominantly again, are IT and
networking teams.
Speaker 1 (35:04):
Did you say dogfooding?
I haven't heard that before.
Yeah, dogfooding.
Speaker 3 (35:09):
So it kind of started with Google a little bit about
15, 20 years ago, and thenothers before it.
So Sun Microsystems had it,silicon Graphics had it, apple
had it for a little bit too.
Which is, our colleagues alltry to run meter networks in
their homes too and with thelatest builds, so that in this
(35:32):
process we're developing it andusing it ourselves.
It's running in our office, inour labs, in our warehouses, all
of that together and it's likepart of this process.
But we, like right now I'mrunning the latest build, I
think, from Friday or Monday atmy house right now, so I get to
see exactly what's happening,what might be broken, what is,
(35:54):
does a feature feel right?
And again, that's reallyimportant too, which I think was
lost in networking too, whichis somehow networking got all
the junk software.
Networking didn't get beauty,and I don't mean beauty as in
just pixels, but actually how itworks, how it functions and how
it flows through.
(36:14):
And we think that part isreally important too, which is
actually all of us using itdeeply, knowing these products,
understand these products and asusers, does the beauty actually
flow through the product?
And yes, it has to lookbeautiful.
But beauty isn't just form,it's also function.
Speaker 2 (36:34):
Yeah, I do have another one and you know, I know
, andy, you said ask hardquestions, so I'm going to.
I don't know if this is goingto be a hard one, but I'm going
to, I'm going to try to go thatroute.
So, and I and I I asked thisquestion out of the context is
I'm someone that works for anetworking vendor.
I know how important trust iswith your customers Because
(36:56):
obviously, like you said, if thenetwork is down, everything's
down right.
So when you want to introduce anew networking vendor into a
customer environment, a hugething is trust.
So another thing that's veryprominent in networking is the
need for integration withsystems, endpoints, et cetera,
right.
So if we think about thiscustomer base where you have,
(37:22):
where you say you're doing, 45%brownfield migrations onto meter
, what does that look like froman integration perspective?
If, say, potential customerslike, look, I really like what
meter's doing, but I, maybe Ilike I either one, I really like
my current wireless vendor, myswitching vendor, and you know I
(37:44):
have licensing that runs outthrough this end of the year and
I don't want to move this dayone.
Is there integration optionsfor that?
Fair enough, easy question.
Speaker 3 (37:55):
We sell it.
We sell a network or we don't.
We'll buy out your licenses.
We'll buy out your hardware.
We'll make it easy on you butwe sell networking.
We don't sell Wi-Fi.
We don't sell routing.
We don't sell switching.
We don't sell security.
We don't sell DNS security.
We don't sell SD-WAN.
We don't sell VPN.
We don't sell ISP management.
(38:16):
We sell networking.
Speaker 2 (38:18):
That's great.
I mean, I actually think that'sa great answer because I think
if you, as the vendor, if youcan stand firmly on what your
product is and what it isn't andyou stick to that, then I think
that goes a long way.
So I actually like that answera lot.
Good, a long way.
(38:38):
So, um, I actually like thatanswer a lot.
Good.
Um, I guess kind of compoundedwith that like what, what do you
tell customers that are kind ofon the fence like hey, I don't
know if I want to completely ripout my stack and put in
something brand?
new right, you know, I mean yeah, I mean, if we use the apple
example, there are there'spluses, there's minuses to kind
of a I won't say a closedecosystem, but you kind of know
where.
I won't say a closed ecosystem,but you kind of know where I'm
going right A full stack that'sdriven by one vendor, right?
(39:01):
So what is your kind of mainselling point that you would
articulate to prospects, newcustomers in that camp?
Speaker 3 (39:08):
So I think it's three things.
It's P P, p, product process,pricing.
First, on the product sideperformance, reliability,
security and integrated products.
I think the delight of beingable to go to our DNS security
and saying you know, andy, can'twatch any more TikTok that
(39:31):
being a single line, it could becrushed.
That being a single line, itcould be crushed.
That being a single thing to doand not having to go buy a
separate DNS product andfiguring out how to integrate it
into the firewall.
And what's going on with theother DNS?
How's it going to propagateproduct at the end of the day?
How well we perform.
If you look at even VPNperformance, we blow away
(39:54):
everybody in the industry.
Straight performance we blowaway everybody in the industry.
Straight performance you knoweverybody talks about is nobody
does benchmarks anymore Because,guess what?
Everybody's asymptotically gonetowards mediocre stuff.
As meter gets more out there,you'll see us actually do
benchmarks.
We believe what we're buildingand we know what we have.
(40:14):
That's much better, so muchbetter reliability, security,
performance, integration andbeautiful products.
That's the first one.
It's on the product side.
Second is on the process thatis delivered because of the
products.
Like I mentioned, right when acustomer signs up.
We have our connect productthat enables us to help them
(40:34):
procure ISPs.
Then we have an entirelyonboarding product where
networking and IT teams can inline, answer a bunch of
questions on how they want thenetwork to be.
All of it, every single pieceof our hardware, down to a port
on a switch, is virtualized inthe backend.
This has become like an overusedterm now digital twin kind of
(40:57):
thing, but in practice I'll tellyou what it means for us, which
is every configuration of anetwork can be actually
configured in the backend, inthe cloud, before it's applied
down to the hardware.
There's no way of bricking thisthing, and so going from ISP
onboarding, configuration,integration and deployment of
(41:19):
networks for a lot of customersprocess is what's attractive.
And the last one is pricing.
We generally in networking it'sbeen huge CapEx and huge OpEx
with licensing and renewals andall these things to purchase
networking gear.
We are zero CapEx and if you dothe TCO out over the course of
(41:45):
three, five, seven years, it isremarkably in the favor of meter
customers.
So depending on who you are,how you're set up and where you
are in the life cycle, it couldbe some combination of product
process pricing.
Speaker 1 (42:02):
Yeah, got another hard question.
Sure it's not really hard andI'll set us up with you know.
So again, I think it'scompelling.
Even the ISP marketplace, likethe first stop that you hit in
this experience I've never seenanything like that.
We at places I've worked, wehad entire teams of analysts
(42:24):
that did all the telco stuff forus because we were busy in the
iron.
So the fact that you can justgo to a screen I thought you
said that it's open to you youdon't have to be a customer to
even you don't have to be acustomer to even you don't have
to be a customer.
Speaker 3 (42:37):
It's what we're providing to the industry.
Please go to metercom slashconnect.
It's called Meter Connect.
Use it, give us feedback.
We'd love to hear from you.
Anybody can use it, send it toyour friends, send it to your
aunts, send it to your kids.
Speaker 1 (42:51):
And there's pricing in there too.
Right Like I can go in and Ineed an MPLS.
Speaker 3 (42:54):
Go from saying here's what I want from 5G coax,
shared fiber, dedicated fiber,starlink.
Here's the pricing.
Just give me a contract thatnormally would take weeks what I
just mentioned Absolutely To beable to do.
(43:18):
So all of that is in there.
Connect is super loved product.
Speaker 1 (43:20):
Yeah, andy, I think your kids would love it too, for
sure, so I can order it rightfrom there, right?
And then does that kick off.
That's right, I mean, I'll getan install date, or right.
Speaker 3 (43:31):
That's right.
Speaker 1 (43:32):
It just happens, that's pretty.
That's right Again.
For anyone who's had to try toorder a circuit, this is very
simplified process.
It's it's a.
It's a logistical nightmare theold school way.
Um, again, hire teams of peopleto do all that and and manage
the billing and forget it rightyeah so.
Speaker 3 (43:52):
so that's for sure.
For one location where youdescribed as soon as you have
multiple locations that havecome up in different times
renewal management, contractmanagement, that stuff is harder
Then.
What you're also talking aboutis I now have 500 locations with
a primary and backup.
That's a thousand ISPconnections I'm managing and I
(44:15):
have to pay.
So Connect has that buying,contracting, consolidated
billing, so you can get a singleline item.
We can pay and make that easierfor you, and because we buy so
many connections, in majority ofthe cases we get you better
pricing than you can ever get.
Speaker 1 (44:35):
That's nice and the billing's integrated right,
Because I can't tell you howmany outages I work because we
forget to pay the bill.
Right, I mean, it's a thing.
Speaker 3 (44:43):
I hear that all the time.
Speaker 2 (44:44):
Yeah, okay.
Speaker 3 (44:46):
So what does?
Speaker 2 (44:46):
support.
Look like in that scenario?
Like, if you know there's acircuit outage from something
that you've procured throughConnect, who do you call?
From something that you'veprocured through Connect, who do
you call?
Speaker 3 (44:54):
So it's two separate models.
If you have our entire network,there's one support.
If you're just buying the ISPusing the marketplace, it's a
different one.
If you have our full network,we have host monitoring built in
.
We have ISP monitoring built in, because it's art of networking
.
I'll give you a preview onwhat's coming.
(45:15):
We will automatically track howyour ISPs are doing against the
SLAs you signed as well.
That's coming.
So if you have the entirenetworking stack from us, we can
keep track of the whole thing,and then we're the ones doing
through our APIs and contacts.
We know when it goes downalerting that comes in.
You get the full service.
(45:35):
But if you just use themarketplace to buy, that's
whoever is on your team, howeveryou want it.
We just made that transactioneasier for you, but we're not
responsible for that circuit.
Speaker 1 (45:48):
Are you monitoring utilization?
Oh yeah, If I'm gettingsaturated and I should upgrade
and all that stuff, Becauseagain that's a whole other thing
, right?
Speaker 3 (45:56):
Oh yeah, we proactively let our customers
know that you're saturating it.
Here's what's happening.
This recently happened with uswith a robotics company Because
of how they were gettingthrottled.
I think we ended up finding abug in the I won't say which
carrier, but in the carrieritself.
Then we were able to do that,but we proactively, a lot of
(46:17):
times let our customers know.
I think the times we'reactually most proud of this
happened recently, when we tellour customers they're paying for
a much bigger circuit than theyneed and say here's all the
data from the last six months ora year You're actually spending
$2,500 on that primary circuit.
You don't need that two gigcircuit, you're okay with one
(46:38):
gig.
Save a thousand dollars a month.
Those are the ones we'reactually the most proud of.
Yes, we tell them when theyshould spend their most money,
but I think it's really nice tobe able to go to a customer and
say please spend less.
Speaker 1 (46:49):
The hard question I never got to.
I was trying to set it up and Idid a poor job doing it, but
this is obviously a lot ofbenefit for businesses, right?
So this is a show consistingmostly of network engineering
folks.
So and you and I had talkedabout this before we started
recording I seem to have how canI say this and not sound like a
(47:16):
big wimp but I becomethreatened by things that
disrupt the model.
Let's say that.
So you know, automation, I felt, was a threat to my job, right,
cloud, I felt, was a threat tomy job.
Now, as it turns out, as I'velearned over time, that stuff's
additive, right, that it's notthat it's taking away my job.
(47:36):
It's that these things likeautomation and cloud can
actually make my job better oreasier or more efficient, or I
can offload things for mycompany in the cloud, but it
doesn't mean they're firingother networking people.
So, to bring this full circle,meter is doing my job, I think,
and I am threatened by it.
So what do we say to networkengineers who look at your full
(48:00):
integrated, you design, youdeploy and you manage the
network, which is, I think, whatmy job is?
What do we say to those folkswho might be like, well, this is
great, but this is a networkengineering show and they're
taking our jobs?
Speaker 3 (48:16):
Yeah, and I think it's a valid question.
You know, andy, you'resomewhere in mid thirties.
You know you've been doingnetworking for the last 15 years
or something.
But I think the reality is thisyou know, I studied networking
about 20 years ago.
Now, when I go back to theuniversity where I studied, they
(48:40):
don't even offer the networkingdegree anymore.
That's just a fact.
As a network engineer, we haveto come to terms with this fact.
I recruit from some of the bestuniversities all the time
Stanford, mit, whatever.
These kids don't even takenetworking classes because
they're not even offered thatmuch anymore.
They get one week of networkingin a distributed systems class.
(49:01):
So I think the first thing toactually realize is the inflow
of talent just isn't there.
That's one.
Two, it's been the same numberof network engineers for a while
.
There's actually a massiveshortage.
If you look at companies likeGoogle that have Google
University and other things,they're trying to train others
(49:24):
into networking because the flowisn't just there.
I'm happy to send you this link, but you should look up Bureau
of Labor Statistics data on iton how many software engineers
have grown network engineerswhich don't have the inflow
that's coming in.
So it's roughly been the samenumber of network engineers for
a while.
But coupled with that, there'stwo things that are happening.
(49:45):
One is the number of devicesand applications on a network
have grown dramatically.
Network have grown dramatically.
Two it used to be that therewas one IT or networking person
for every 75 to 100 people in acompany.
Today, the average is about 250.
So you are in a situation wherethere's no new inflow of talent
(50:08):
and the existing people askedto do way more than they
currently are.
And the network has become moreimportant than it's ever been
because everything in the worldis packets.
Now, as we all get old andactually studied networking, I
don't know who's coming behindus, but even the folks that are
today doing it.
(50:29):
If you ask them hey, do youhave time for that project you
want to get to, or anything, orhow are your KPIs?
How are you dealing with it?
When was the last time youmonitored the network?
When was the last time you didupgrades?
It's a lot to be able to dowith no new talent coming in,
more expectations and morenetwork load than ever before.
Speaker 1 (50:50):
It's untenable, right Like we seem to be heading
toward catastrophe.
Speaker 3 (50:55):
So that's actually another huge premise of why we
started Meter is, if you're inthat position, we did not feel
like there was a company thatwere building hardware and
software for me as a networkengineer, for me to be able to
do my job well, which is servemy colleagues so that the
overall company can do betterNetworking is this Archimedes
(51:15):
lever for the rest of thebusiness?
One of the things I really careabout is making networking cool
again, like truly, but I dothink companies like us.
We hope to be augmentingnetwork engineers so that they
can do much better things fortheir customers, but the reality
(51:36):
also is there's not enough ofus.
Networks are falling apart.
There's end of life, end ofsupport networks out there, with
some of the largest retailerswith credit card data flowing
through that who knows when wasthe last time they did anything
with it.
So I hope what we actually dois build such great hardware and
(51:56):
incredible software.
That's why even something likeCommand came in.
Is that we want to buildsomething so good that all the
folks that are young, like you,andy, that grew up doing CLI,
would love to use Command.
Speaker 1 (52:09):
Just keep calling me young man.
No, that makes a lot of senseand it rings true to a lot of
the things I've been hearing forat least the past year or two.
We don't have new talent comingin.
I can't believe.
Yet I keep hearing that thatcomputer science courses have
(52:30):
basically networking from thecurriculum.
You said something earlier onthat rang true for me when I was
a cable guy at an ISP and I sawthe statistics, the you know.
You know today we have fivewire.
You know five computing devicesin a home and you know, in two
years we'll have 10.
And there was kind of thisMoore's law type math of the
(52:52):
explosion of devices just inresidential areas.
And that's when I thought, wow,networking will grow for quite
some time and I will definitelyknow this is a good career to go
into.
So it's such a strange thing tohave that you said it before
that explosion of end users andeverything is on the network.
(53:13):
I mean we have thermometers andfish tanks now that are IoT.
There's just so many devicesand the fact that we don't have
the talent coming into theindustry to support all this and
it's just going to keep growing.
So I'm comforted that a companylike yourself is addressing that
?
Because it sounds to me like ifI am a network engineer working
(53:36):
at a company using meter, I cando a lot more with less people
like myself.
I mean, we have less of us.
So I guess we need tooling likemeters to be able to manage a
huge infrastructure, get moreinsights, that more kind of
integrated solution.
So isn't it funny I'mthreatened by meter and yet
(53:58):
you're trying to help me.
So I'm glad we're having theseconversations out loud, because
when I see this I'm like, ohgreat, as a service, they're
trying to make me go away, butthere's less of me every day,
every month, every year.
Like we need help.
Speaker 3 (54:11):
Yeah, I mean, we think of network engineers as
farmers and we're just shovelsImagine doing farming without
shoveling or a tractor or a plowand we felt like you know,
there were some legacy vendorsthat build a shovel, a plow but
it kind of stopped there.
We need tractors for networkengineers and it needs to be
(54:34):
really great products.
Again, the simplest way for youto know if people have cared
about networking the decade ortwo is if you pick up a
networking product that might behardware or software, do you
actually find it beautiful?
And I'm sorry there's not asingle beautiful networking
product out there until we camealong.
That tells you how much peopleactually care and we want to
(54:59):
build that.
That lever is important.
That productivity multiplier isimportant.
Speaker 1 (55:03):
And this has been very informative for me.
We got some great interactionsin the chat.
I think we found you a DAScustomer, so that is awesome.
Any parting thoughts?
I think we covered.
You know all the questions Ihad and then some.
I think what Meter has iscompelling.
(55:24):
It solves a lot of problems fora lot of different use cases.
Anything before we close thisout.
Speaker 3 (55:30):
Yeah, I think three things that I'll mention.
One we really care aboutnetworking.
There should be great hardwareand software.
Hardware should not becommoditized.
Software should not be ugly.
Two, we do everything from ISP,routing, switching, wireless
DNS security, sd-wan, vpn, allintegrated.
(55:53):
That's the way it's supposed tobe.
And three, I love networking.
So if anybody wants to reachout, my email is anil at
metercom M-E-T-E-Rcom.
I'll respond pretty fast.
Speaker 1 (56:07):
Awesome.
Thank you so much, anil.
This has been a great session.
Chris, thanks so much forjoining.
If you want to check out Meter,you have Anil's email address,
which is pretty cool.
I don't have many emailaddresses of CEOs.
You're going right in mycontacts.
I'm going to bug the heck outof you.
You can also find them atmetercom.
They have that awesome freeresource which I'm trying to
(56:29):
find here.
It's metercom.
Speaker 3 (56:31):
Where is it?
Slash connect.
Speaker 1 (56:33):
Slash connect, you can check out their marketplace
of all kinds of ISP connectivity.
I love it.
And to check out everything Artof Network Engineering, you can
check out our link tree atlinktreecom.
Forward slash, art of NetEng.
Great conversation.
Thanks so much, anil Chris, andwe'll see you next time on the
Art of Network Engineeringpodcast.
Engineering Podcast.
(57:00):
Hey everyone, this is Andy.
If you like what you heardtoday, then please subscribe to
our podcast and your favoritepodcatcher.
Click that bell icon to getnotified of all of our future
episodes.
Also, follow us on Twitter andInstagram.
We are at Art of NetEng, that'sArt of N-E-T-E-N-G.
You can also find us on the webat artofnetworkengineeringcom,
(57:21):
where we post all of our shownotes, blog articles and general
networking nerdery.
You can also see our prettyfaces on our YouTube channel
named the Art of NetworkEngineering.
Thanks for listening.
This is the Art of.
Speaker 2 (00:01):
Network Engineering podcast.
In this podcast, we'll exploretools, technologies and talented
people.
We aim to bring you informationthat will expand your skill
sets and toolbox and share thestories of fellow network
engineers.
Speaker 1 (00:23):
Welcome to the Art of Network Engineering.
My name is Andy Laptev andtonight we have a very special
episode.
This is a sponsored episodewith the good folks at Meter.
Meter is a networking as aservice company.
They are making really coolstrides in the NAS business and
(00:44):
we're going to do a few thingslike define what that is and see
what kind of problems they'resolving and how they're doing it
.
Anil is a really cool guy.
He's a career network engineer.
He's one of us people, so Ithink that that's really
important.
If you're going to come on theshow, you better be a network
engineer talking about yourproduct.
So Anil has the credentials.
(01:04):
We're also joined by ChrisMiles in the future in Australia
.
How are you doing, chris?
Speaker 2 (01:08):
Doing well.
Brother, Thanks for having me.
Speaker 1 (01:09):
Absolutely.
Thanks for coming on.
So, anil, let's tell folksabout yourself.
Who are you, what do you do?
How did you get to be a networkengineer to this CEO of a NAS
company?
Speaker 3 (01:25):
Andy, chris, thanks for having me and thanks
everybody for tuning in.
I studied networking in collegeand I've always felt networking
is one of the most importantthings in the world.
And maybe a career bet mybrother and I who I started
meter with made about decade anda half, two decades is that we
all will use internet more thanwe currently do, and we thought
(01:48):
that was a pretty safe bet tomake at that time and we
continue to believe that's avery safe bet to make today.
And everything that we do thispodcast or doing anything it's
all just packets somewherefloating around through wires
between radios, and we wanted tohave an opportunity to build
(02:11):
really great hardware, reallygreat software for networking,
and so meter at the core is justa networking.
We believe in a couple ofthings that I think the industry
lost a little bit of track of,especially as users and
customers.
We felt like everybody thoughthardware was commoditized,
nobody was really payingattention to software and nobody
(02:33):
was integrating productstogether.
So what we set out to do is,instead of building a point
solution, instead of sayingwe'll just build access points,
somebody else will figure outthe rest, we'll just build a
security product, somebody elsewill figure out the rest.
We'll just build a securityproduct.
Somebody else will figure outthe rest.
We said why don't we take thetime and build really good
products for the entire stack?
And so today, meter buildseverything from routing,
(02:56):
switching, wireless PDUs, ispmarketplace that's built in
SD-WAN, vpn, all of it wrappedtogether along with applications
and software to give networkengineers really good control
and really good visibility tothe level that I would want it
(03:18):
as a network engineer.
You mentioned networking as aservice.
We don't really think ofourselves as a networking as a
service company.
Oh, I apologize.
No, no, you don't need toapologize.
Networking as a service issomething we do, but at the core
, what we really believe is onnetworking itself across
(03:38):
hardware, software, operations,and the four big product lines
that we have that do this is.
One is our ISP product that'scalled Connect.
Think of it as Kayak or Expediafor ISPs.
How it started is that we foundit really intriguing that you
couldn't buy the internet on theinternet, so that's how it
(04:01):
started.
Then we have our wired productsthat are routing, switching.
Then we have our Wi-Fi productand about a month ago we
announced our cellular productand what we're hoping is in any
space if there's a packet moving, it goes through meter,
hardware and software and reallyfrom designing the hardware
(04:32):
operating systems, firmware,distributed systems, apis,
applications we just think itcan be way better If there's a
company that really cares aboutthe hardware, cares about the
software and integrating it alltogether.
We got started about a decadeago.
We've been relatively quietbecause we wanted to start
talking about meter when we knewwe had the best products for
networking.
So that's why you probablyheard more about us in the last
(04:52):
couple of years, in particularin the last year, because we
feel like we've crossed thatthreshold.
So that's who we are, that'swhat we do.
Speaker 1 (04:59):
That's a huge like you've integrated the entire
stack.
That is a big.
That is a big task.
You are you are you are verybrave and I don't know why, as
you were talking, I kind ofthought of just you know, I'm
trying to think of otherintegrated solutions and, like
Apple came to mind, right, theyhave the hardware and the
software and they're able toprovide a more integrated, more
(05:20):
secure, better experience than abunch of disparate parts and it
kind of sounds like that'swhere you're going is it's?
it's your hardware, it's yourNOS, it's your people, it's it
sounds like a good.
Give a much, much cleaner, um,I guess, experience for for your
customers.
So so who?
Who are your customers?
(05:41):
Right, like I'm a networkengineer.
So is this something?
I mean I guess you're marketingto businesses, correct?
People who need networking,infrastructure and things
installed.
Right, how does someone like me?
Right, it's all about me.
I want to know how this isimpacting our network engineers.
So we're going to get into theproduct.
I like I know that you have.
(06:04):
What's the new?
You have a really cool naturallanguage feature.
Yeah, command, which I'm reallyinto, because the more I see
that in networking platforms, Iget really excited because, you
know, as fun as it is learningarcane syntax in a bunch of
different CLIs, right, which wasthe whole job for a really long
(06:24):
time I love the fact that I canjust speak to the
infrastructure in naturallanguage and get insights.
I think that that's supervaluable.
You know, there's a lot ofAI-ification of everything,
right, and we've almost becomekind of immune to it, like oh,
great AI.
But when I saw the demo up onyour site, I'm like, oh, this is
(06:51):
like, this is real, this isuseful, I can ask a question and
get really valuable insights.
It looked like I could build, Ithink, like my own dashboards
too, based on the data that.
So I think I know this is amainly audio podcast and I think
that we should show some stuffat some point, you know, during
this.
I know it's still early.
I'd like to see, but I guessback to my original question.
So let's say, I'm a businessand I'm starting up and I need
internet, right, I needinfrastructure.
(07:11):
So I call you guys and what doyou do for me?
I think you do everything.
Speaker 3 (07:16):
Yes, so our customers are folks like you mentioned
that are starting up newbusinesses, all the way up to
some of the largest companies inthe world, so large public
companies.
A lot of network engineers arelistening when we all congregate
on the internet.
If you imagine the form of theinternet that runs on meter.
(07:39):
The largest hedge fund in theworld, bridgewater runs on meter
.
Whole gamut of companies.
Actually, chris works at acompany that runs on meter,
another networking company thatI just found out before we hit
record and so our customers andwho we're building for is
entirely for network engineersnetwork engineers and the
(08:08):
premise of our business when westarted it was precisely that
nobody was building products fornetwork engineers anymore and
that felt very obtuse.
We felt that as users ourselves.
It felt like random products,random features.
You know forgive me, andy, butcrazy marketing stuff that
didn't make any sense and we'rejust like where are the real
(08:29):
products?
Who are the people that reallylove this?
So our products are built fornetwork engineers.
Our customers range frombusinesses that are just getting
started up to the largestbusinesses in the world that
have the most security,compliance requirements and
really large technology andnetworking teams, and when
(08:50):
somebody signs up.
Our work starts with firstmaking sure the right ISP gets
to this business, because nomatter how great of a network
you have inside of a space, ifyou don't have great when coming
in and out, kind of pointless.
So that's where our Connectproduct comes in, which is
exactly that of Kayak or Expedia.
You just literally type in anaddress, we'll show all the ISPs
(09:14):
you can purchase and you canpurchase it instead of waiting
two weeks of going back andforth to figure out who's
available and what price.
And we actually made thatentirely a free product because
we wanted to also contribute tothe community.
So anybody can go to metercomslash connect and just use it.
Then we work with our partnersso some of the largest channel
(09:37):
partners to some of the smallestones or us that understand a
space, and we design a network.
So both we can work with thechannel or work directly with
the customer to design a network.
Then it's our rack that'sinstalled entirely, that has
PDUs, routing, switching andeverything you can imagine there
(09:59):
, from high availability to hostmonitoring and DNS security and
stateful firewall and all thosethings that are built in that
you would expect that are deeplyintegrated together, along with
indoor and outdoor wireless.
And then now, as I mentioned, acellular product that, instead
of you having to go spend a lotof money on DAS and other
(10:21):
systems, cellular deployed justlike Wi-Fi, we deploy that.
Then we give what we hope isbeautiful software with our
dashboard and things likecommand that you mentioned, that
give full control to ourcustomers, who are, again,
primarily the networking teamsat the businesses that we work
(10:42):
with.
Speaker 1 (10:42):
So quick question is this how do I interact with the
infrastructure?
Is this a GUI?
I'm guessing I have an API onthe backend.
Speaker 3 (10:50):
Yeah, just primarily the dashboard and primarily
command.
Command.
Like you mentioned, we launchedit about two months ago and
it's been an incredible success,not just in networking but in
technology as a whole.
Many people thought it was oneof the best applications and
models that people had ever seen.
So real software directlyavailable to customers.
Speaker 1 (11:14):
So we don't have to learn a new CLI, right?
Yeah, yeah.
Speaker 3 (11:17):
No, yeah, that's where command comes in.
It's beautiful, and one of thethings about software that's
challenging is either you'redesigning it for the novice or
you're designing for the expert.
When you start a new softwareproject, you're kind of told
pick one.
And I think what's beeninteresting about command that
(11:37):
we built entirely in-house it'snot APIs, we're calling
elsewhere and it doesn't go tosomething like OpenAI or it's
all meter is command is flexibleand kind of goes around,
depending on the user, and thepremise of command is that
software should be soft, and so,along with the dashboard that's
(11:59):
fully fleshed out, command cando things like give information
for anything you're asking, takeaction.
It can be really complicatedaction.
Like you know, create me a newSSID with this VLAN and make
sure it auto rotates every dayat 2pm and put it on all the
access points.
You can actually say that andthen we'll go do it, and then
(12:20):
command also just generatessoftware on the fly, just the
way you want it.
We want every one of ournetwork engineering and IT teams
to have a software engineerjust for them, but it's a full
software platform integrateddeeply into the hardware.
Speaker 1 (12:37):
So one of the reasons I love live streaming is we
have a live audience and I mighthave just found you a customer
because we have someone in thechat saying wait, did I hear
right A DAS replacement?
I might be interested.
We're looking for a newsolution.
I am not familiar with DAS.
I'm Googling it on the side sowithout me giving a terrible
(12:58):
explanation?
Do you want to hit a high level, like what is DAS, and why
would this gentleman be smart ingetting it to you, sure.
Speaker 3 (13:04):
So DAS stands for Distributed Antenna System.
So when you go inside of abuilding today, when you open up
your phone, the goal is bothfor the carriers and IT teams is
that you should have five barson your cell phone.
But that's actually really hard, challenging and expensive to
do On the carrier side.
Carriers have to go installantennas, rent out rooftops and
(13:27):
spend a lot of money, and thencompanies install these systems
called DAS systems.
If you pull up an image, go toGoogle Images, you'll see there
are these antiquated systemsthat look like servers in Matrix
that go in basements and thenthey have antennas that are
distributed throughout abuilding and so when you pick up
(13:49):
your phone, your phone is goingfrom you know signal from your
phone to one of these DASantennas to one of these outdoor
antennas that are popped up,and then it just meshes all the
way through next stop, next stop, next stop to get to whichever
carrier that you are.
What meter cellular enables youto do is a couple of really
(14:10):
interesting things.
Just like there's a Wi-Fi AP,there's now just a cellular AP
and it's connected to the fullnetworking stack that we built.
This is the advantage ofbuilding integrated products.
And now, when you open up yourphone, instead of it going from
your phone to an indoor antenna,then hopping all these external
antennas, it actually just goesfrom your phone to the cellular
(14:33):
AP and directly throughinternet to whichever carrier
you have.
So you, as a customer, can getgreat cellular that's deployed
in 75% less time, with somethinglike 80% less of the cost, and
cellular deployed as easy asWi-Fi that's again integrated
(14:54):
into the entire networking stackthat a company already needs.
Speaker 1 (14:58):
All right, so hopefully our listener Nully
heard that.
They should reach out to thesefolks.
It sounds like they have areally good solution.
Speaker 3 (15:07):
Yeah, shoot me an email, Anilatmetercom.
I'm super simple.
Speaker 1 (15:11):
I never thought about , I guess, cellular in large
buildings and how difficult thatwould be.
I mean, I know how hard it isto get Wi-Fi in buildings.
But yeah, the cell towers areoutside.
You know, you don't havetypically access points inside.
Speaker 2 (15:22):
So if you see those little knobs on the ceiling that
aren't APs.
That's typically what those are.
Speaker 1 (15:27):
They're little antenna repeaters basically.
Speaker 2 (15:31):
I do have a question Now.
You've kind of already touchedon this and I think I know where
you're going, but I want togive you the opportunity to
expand on it.
So you know, if I think aboutthe landscape of networking
vendors today, obviously there'sa lot of them, or there's a lot
out there, there's a lot thathave been around for a very long
time and you know they probablyhave claims as well that they
(15:53):
do everything in the stack, kindof similar to what we're
talking about today.
But my question is you know, ifI think about your all stance,
obviously you're relatively newto the game.
You've already made quite a bitof headway.
Your all stance, obviouslyyou've, you've, you're
relatively new to the game.
You've, you've, um already madequite a bit of headway.
Do you feel that being astartup in this scenario kind of
helps you guys with umreleasing better products, just
(16:13):
because you don't have the yearsof technical debt that maybe
some of the other people do inthis space?
And can you maybe elaborate onon what your, what your uh input
is there?
Speaker 3 (16:23):
yeah, we've only been around 10 years where, like you
mentioned, there's others thathave been around 20, 30, 40
years.
So totally we've only beenaround 10.
We've not been around as longas some of the other folks.
I think it's less technicaldebt because of the years.
(16:45):
I actually think it's moreabout our business model and how
our incentives are aligned withour customers that enable us to
do things that others can't.
So a couple of things that I'llmention here what that means.
One is meter builds an entirelyintegrated stack.
We don't sell individual piecesof hardware, so we don't just
(17:06):
sell an AP.
We don't sell individual piecesof hardware, so we don't just
sell an AP.
We don't just sell a switch.
We sell networking.
What that enables us to do isalways know what's above us in a
stack and what's below us andhave the right APIs, right
protocols and right thetransactions between each piece
of hardware to be able to buildgreat networks.
It enables us to make differentchoices on hardware.
(17:28):
We are not expecting our APs toalso do some routing and some
switching.
Our APs only do wireless andthat leaves a lot of room on the
chipsets, because we all usethe same Qualcomm chipsets.
We use the latest, greatestQualcomm.
So does everybody else.
But it allows room to go domore and really focus on
(17:50):
wireless really focus onwireless.
And our business model also isthat we take on the capital risk
entirely instead of ourcustomers.
What that means is there's zerocost for the hardware,
including when we release newhardware.
So every four or five yearsthere's a new Wi-Fi spec for my
triple E, or when there's newswitching algorithms or routing
(18:12):
or new hardware, we make newhardware.
Or when there's new switchingalgorithms or routing or new
hardware, we make new hardware.
All of our customers getupgraded to that new hardware as
well.
But what that means with ourbusiness model that we take on
the capital risk and hardwarecosts are zero is that we only
get paid if we provide a goodproduct and a good service to
our customers.
That means our incentives withour customers are very much
(18:32):
aligned.
We are not looking to make abox and sell that for 25% margin
and hope to sell you a new boxevery 18 months.
25 is low, I was being generous.
But our business model isentirely aligned with our
customers.
And if you take a step back andthink about it, networking is
(18:55):
the last part of the technologystack that, until meter came
along, wasn't aligned withcustomers.
Everything from compute tostorage, to CRMs, to everything,
was that if you provide me agreat product, that's when you
get paid.
But networking was that you payme upfront.
And great product that's whenyou get paid, but networking was
that you pay me upfront andthen you'll never hear from me
(19:17):
again until I want to sell youanother thing.
So that actually enables us todo much different things on how
we can integrate hardware,integrate software and build
products other people can't.
Because we know what's above usand below us in the stack,
because each part of our stackis built to do one thing that
(19:37):
it's good at and great at, andhow our business model is set up
so that our incentives arealigned with our customers.
Speaker 1 (19:46):
Yeah when we were originally talking.
We did an episode a while backon networking as a service and
somebody in the chat saidnetwork attached storage, no,
the other NAS.
But I wondered how?
Because hardware is soexpensive at least traditional
hardware that we're used to Ihad thought how do you get a
(20:07):
return?
Like for you to just hand metens of thousands of dollars of
hardware.
I'm assuming I'm signing acontract because you need to get
a return, I guess, on all thehardware you just handed me.
So is it safe to say that we'rein some type of contract so
that you can get your return?
Speaker 3 (20:24):
Yeah, not just tens of thousands of dollars.
With our largest customers thathave tons of locations, square
footage and a lot of footprint,millions and tens of millions of
dollars of hardware, and Idon't mean to make it sound
different than the traditionalmodel, because you're in the
same boat.
Speaker 1 (20:40):
If you spend $15 million with a big vendor, right
?
You're in it for five to sevenyears, and whether you have a
good experience or not.
So it's not a downside of this,but I'm just wondering.
I'm expecting some type oflong-term relationship, just so
you can it is long-term Three,five, seven years are generally
what we see from customers.
Speaker 3 (20:59):
But what's implicit in all of these contracts that
we do with our customers is thatwe're providing them great
products and great services, andthat's where it's different
than before, where you're notjust giving us the money and
then that's the end of it.
We have to earn the right tohave you as a customer, but it
is definitely something becausewe believe in building the right
(21:21):
products.
It's definitely a bet we'remaking that we will get to keep
the companies as customers, butalso those networking teams, no
matter where they go, ascustomers as well, because we
would like to obviously buildthe largest networking company
possible, and not only do wewant to keep the companies, but
also the people forever.
Speaker 1 (21:41):
How do we get you deployed?
Is this Greenfield Good?
Speaker 3 (21:45):
question.
About 40, 45% of what we do isBrownfield, and what we also
have is a buyback program, sothere's no sunk costs.
Let me break that down a littlebit Generally.
When an IT team has to make adecision, they have to work with
finance, they have to work withprocurement, they have to work
(22:05):
with vendor management offices.
It's not just a technicaldecision, it's a business
decision, and they have othercolleagues and departments that
they owe an explanation to whichis hey, we just spent $10
million on networking gear.
What's happening?
You want to use this new thing?
What happened to all that moneythat we just spent?
Or you just spent 100K.
What happened to it?
(22:26):
So we actually buy out all thelegacy hardware that companies
have to make the switching costseasier over the meter as well,
so that it's much more palatablefor procurement, finance and
other teams.
And again, we serve IT andnetworking teams and our job is
to make their lives easier.
Speaker 1 (22:47):
I'm guessing that you're also.
You're attractive to thefinance departments, right is I
know?
That a lot of the as a servicestuff is, you know, opex versus
capex, right?
So that's?
I mean that's this model right,I have a recurring bill we have
both models some cost, oh okaywe have both models.
Speaker 3 (23:04):
Um, you know, we have customers that are owned by pe
firms.
Opex hurts their eDA margins,so we have capital lease models
for those folks as well, and wehave pure OPEX models as well.
But what's inherent in both isthat we're taking the capital
risk on the hardware, not ourcustomers.
So our customers get to havethe benefit of the fact that
(23:29):
Meter is taking the risk on thehardware, that meter is taking
the risk on the hardware andmeter is on the hook to provide
really great products andservices to customers to earn
and continue to earn theirbusiness.
Speaker 1 (23:40):
You told me to ask hard questions, so I'm going to
Please Otherwise it's not funfor me either.
Well, no, but I think it's fair.
So I'm following you and so farthis is compelling.
I have experienced qualityissues right With different
vendors software, hardware.
So how's your quality, if youcan make a generalized statement
(24:02):
?
You know, I'm guessing you havesupport for RMAs, right.
Hardware is going to fail.
Nothing's perfect Because it'sintegrated.
I'm assuming it's a very tightintegrated experience and the
assumption I'm making is thatthe quality is high because I've
seen your customer list.
It's large.
If you weren't doing a good job, people would be yelling about
(24:23):
it on the internet, becausethat's what people do on the
internet is they yell ateverything they're mad about,
and I've never heard or seenanyone yell about how mad they
are at meters.
So can you talk about yourquality?
Just your NAS, your hardware,things that would be important
for folks.
Speaker 3 (24:36):
Yeah, I think.
Before jumping into specifics,which I will, I want to tie it
back to what I was mentioning onthe business model itself.
What's tied to quality isactually speed to market, and
one of the challenges we thinkthe legacy vendors have with
their model is I'm sure you'veseen this in marketing from all
(24:57):
these companies they want to andneed to say world's first Wi-Fi
7 AP.
They all want to and need toAbsolutely For a lot of
different reasons.
They need to be first to market.
They're like we're doing Wi-Fi7.
I'm going to put this randomchip in there, this random
protocol, just because I'mtrying to get Andy to buy one
(25:19):
more time, even though he justbought three years ago.
I need to convince him that heneeds this new box.
Speaker 1 (25:25):
It's that new iPhone.
Every two years I need the newone.
That's right.
Speaker 3 (25:29):
We don't have that affliction.
You will never see us beworld's first.
In fact, those are kind of thesilly statements you will never
hear me to say like world'sfirst Because it doesn't matter
for world's first, does itactually work?
Does it work for me?
So again, it ties back to ourbusiness model is we don't have
to and we don't need to go dothose silly games and play
(25:52):
finite games.
We like to play infinite gamesthat keep going forever.
Second is, like you're saying,because we have this business
model and this verticalintegration and we sell the
entire network, there's threeimportant things that affect
(26:14):
quality.
One is a unified operatingsystem across the entire
platforms is a unified operatingsystem across the entire
platforms and what that buys usis the ability to know that, oh,
ap version this and firmwareversion this for firewall
doesn't work or some feature.
Those issues don't occur.
Actually, when people thinkabout quality issues, they go to
code quality and testingquality issues.
Yes, those things happen, butin reality, the actual issues in
(26:38):
networking for quality comefrom integration issues.
Even with the same vendor,which is you might have this
version of the switch not workwith this version of AP, not
work with this version offirewall and this one feature
wasn't tested properly becauseof that, those kinds of things.
Because we have a single unifiedfirmware image and operating
(26:59):
system, we know what actuallyworks together and those are
releases that are cut that wayevery quarter with that in mind.
What it also buys us is asingle API across the entire
stack, and how that shows up isboth in quality but also in
usage.
When you want to make changes,you can make change in one part
(27:23):
of the stack and actually haveit traverse everywhere, really
simply, and enable us to buildgreat software for our customers
because of it.
And the third thing it buys usis also a unified stats pipeline
actually knowing how the entirenetwork is doing throughout the
pipeline of a packet, from ISP,routing, switching, wireless
(27:47):
through firewall, through DNSsecurity, through VPNs, through
SD-WAN.
All of it is connected with aunified stats pipeline.
That actually was enable us tobuild much higher quality
products because of the stancewe took on the entire network,
because of our business modeland how we're set up to build
hardware and software.
Speaker 1 (28:09):
You just made me think of troubleshooting.
So I have a problem.
People are calling in, theycan't reach an application.
Let's say I'm thinking thetraditional troubleshooting I
would do.
I do some captures, some PCAPs,right.
So how do I troubleshoot this?
Is it any different From whatyou just described?
It sounds like I have insightsthat's right, Just at my
(28:30):
fingertips.
Speaker 3 (28:30):
Yeah, so you can do all the traditional stuff from
understanding everything fromclient health metrics to packet
captures.
But it's a unified thinginstead of you having to go look
at every part of the stack.
It's a huge advantage beingable to get that.
That's a push from the systemrather than a pull from every
(28:51):
part from you as an engineer.
Second, you referenced commandearlier.
One of the things you can ask.
Command is just like Slack orTeams.
You can actually just hit atand get to a location or get to
a meter hardware or get to aclient device.
So you can just hit at and sayAndy's Windows let's say that's
(29:12):
your computer's name and saytell me the health of this
client and give me a five-bulletsummary.
And at the speed of a Googlesearch it will go and summarize
all the things which you can golook at yourself, because you
have logs available andconnection history and all those
things available, and just gethere's the five bullets.
(29:34):
Hey, his snr kind of sucks, oryou know, andy wasn't supposed
to really use 2.4, but he'sactually on 2.4, um, those kind
of things old school man 2.4forever.
Speaker 2 (29:45):
So it sounds like something andy would do,
probably.
How dare you five five?
Speaker 1 (29:50):
gigs, five gigs, forever.
I.
The more you talk, the morequestions I come up with.
So, please, you know we welland Chris jump in, jump in
whenever.
So, can I, can you do cloudpeering?
So, like somebody in the chatsaid, you know they're curious
if BGP supported it.
So that just made me think,okay, well, can I, can I peer to
(30:10):
my CSPs from your infra?
And then what does that looklike?
You know, A, can we connect?
And then B, are there anyinsights there that you can pull
in?
Like again, something's broken?
I'm going back to my productiondays.
Yeah, I railed, I cried against, you know, about cloud, because
when something that was hostedin the cloud broke, it was very
(30:33):
confusing and complicated tofigure out which cloud it's in.
How is that routing built?
What exactly broke?
It was just magic somewhereelse.
So I guess we'll take it one ata time.
Do you?
Can I peer with CSPs withpublic cloud with your?
Speaker 3 (30:47):
Yeah, so we can.
We can connect to a pops publiccloud, direct connect, all
those things.
But I want to also answer thesecond question you had, which
is do we have visibility andobservability and what's
happening there Once you doconnect?
We have it on the connectionsitself, but not beyond.
Those are things we will buildlater on, but we don't have that
(31:07):
today.
I want to be clear that I'm not.
Speaker 1 (31:09):
And I wouldn't expect it to.
That's not even like a fair ask, but I'm just thinking all the
things you connect to right.
Yes, so you connect to cloud,which is good, that's important.
I mean, you guys have thoughtall this through.
It's not like, oh well, we onlydo Wi-Fi in-house and you got
to figure the rest out like youguys.
Speaker 3 (31:23):
No, no, routing, switching, security ISPs.
Speaker 1 (31:30):
And I like that.
It's I guess it's one bill.
I don't care when the billcomes, I'm assuming monthly, not
that it matters, but where I'mgoing with this is licensing.
One of the things that hasalways really made me nuts is,
you know how do I say this?
Subscription licensing.
I don't want to say anybody's,you know patented license term
and then Make it look like I'mtalking trash on that particular
(31:52):
vendor because I'm not.
But the subscription licensingmodel as a person who came up
again it's.
It's tough being around andbeing an old school person,
because back in my day this ishow it was done and it was
better, you know, when I had aperpetual license and that thing
was good forever.
I really enjoyed that, becauseI've been hit with outages when
subscription licenses expiredand we weren't paying attention
to the logs and bad thingshappen.
Speaker 2 (32:14):
So but I'm guessing you're subscribed.
The phrase I heard a lot wasone throat to choke.
I don't know if that's reallywhere you want to go with that,
but I heard that a lot Totally.
Speaker 1 (32:29):
But I don't get the sense from Meter that you know
like.
Licensing isn't going to be aproblem because you're not
hitting me with this separatemonthly bill to keep everything
working and the features up.
It's all integrated, it's onebill everything you need.
Yeah, which I really like.
That's even just simplifyingthe operational model.
Just that alone.
I know it sounds it should betable stakes, like your
licensing model shouldn't bringinfrastructure to its knees and
(32:51):
people at vendors will say, ohno, it doesn't we send log
messages and warnings.
And people at vendors will say,oh no, it doesn't.
We send log messages andwarnings.
And well, okay, yeah, becausewe're all sitting around
watching logs for expiringsubscription licenses, like
right, we're understaffed, we'renot, we're doing other things,
so okay, so, licensing is easy.
It's integrated Code.
Upgrades is another thing thatcame up for me, did you?
Say you release quarterly F.
Speaker 3 (33:11):
Firmware is quarterly , for dashboards and other
things weekly, sometimes daily.
We have a really nice velocity,but in general it's more driven
by customer outcomes.
One of the things we reallycare about which I think also
the industry has kind of lost alittle bit is, as we started
(33:33):
doing more fancier and fancierfeatures.
I think the basic networkingstuff wasn't done to perfection.
So that's the other thing wetry to do is we try to do the
basics really right, like mybrother who I started meter with
.
If you guys have them on andsay like, hey, tell me about STP
(33:54):
and how it should beimplemented, not used, be
implemented, not used,implemented That'll be a fun
conversation.
So we try to do the basics right.
The software is derived fromthe outcomes from customers and
really just going everythingderiving from those interactions
with customers and howcustomers get to use our
(34:16):
hardware and our software.
Because at the end of the day,I think IT and networking teams
have one of the toughest jobs inall of any industries actually,
because they're only thought ofwhen something doesn't work.
When the network is working,nobody's calling a networking
guy or gal saying, hey, thenetwork is fantastic, good job.
(34:38):
Never, it's never happened.
It's only when it doesn't work.
Speaker 2 (34:45):
Proving innocence rather than that's right
Function.
Speaker 3 (34:48):
That's right.
That's right.
So all the software that webuild code upgrades, how we do
firmware, how we do testing, howwe do firmware, how we do
testing, how we do dogfooding,how we do canary upgrades.
All of that is tied back totrying to understand the lives
of our customers, whichpredominantly again, are IT and
networking teams.
Speaker 1 (35:04):
Did you say dogfooding?
I haven't heard that before.
Yeah, dogfooding.
Speaker 3 (35:09):
So it kind of started with Google a little bit about
15, 20 years ago, and thenothers before it.
So Sun Microsystems had it,silicon Graphics had it, apple
had it for a little bit too.
Which is, our colleagues alltry to run meter networks in
their homes too and with thelatest builds, so that in this
(35:32):
process we're developing it andusing it ourselves.
It's running in our office, inour labs, in our warehouses, all
of that together and it's likepart of this process.
But we, like right now I'mrunning the latest build, I
think, from Friday or Monday atmy house right now, so I get to
see exactly what's happening,what might be broken, what is,
(35:54):
does a feature feel right?
And again, that's reallyimportant too, which I think was
lost in networking too, whichis somehow networking got all
the junk software.
Networking didn't get beauty,and I don't mean beauty as in
just pixels, but actually how itworks, how it functions and how
it flows through.
(36:14):
And we think that part isreally important too, which is
actually all of us using itdeeply, knowing these products,
understand these products and asusers, does the beauty actually
flow through the product?
And yes, it has to lookbeautiful.
But beauty isn't just form,it's also function.
Speaker 2 (36:34):
Yeah, I do have another one and you know, I know
, andy, you said ask hardquestions, so I'm going to.
I don't know if this is goingto be a hard one, but I'm going
to, I'm going to try to go thatroute.
So, and I and I I asked thisquestion out of the context is
I'm someone that works for anetworking vendor.
I know how important trust iswith your customers Because
(36:56):
obviously, like you said, if thenetwork is down, everything's
down right.
So when you want to introduce anew networking vendor into a
customer environment, a hugething is trust.
So another thing that's veryprominent in networking is the
need for integration withsystems, endpoints, et cetera,
right.
So if we think about thiscustomer base where you have,
(37:22):
where you say you're doing, 45%brownfield migrations onto meter
, what does that look like froman integration perspective?
If, say, potential customerslike, look, I really like what
meter's doing, but I, maybe Ilike I either one, I really like
my current wireless vendor, myswitching vendor, and you know I
(37:44):
have licensing that runs outthrough this end of the year and
I don't want to move this dayone.
Is there integration optionsfor that?
Fair enough, easy question.
Speaker 3 (37:55):
We sell it.
We sell a network or we don't.
We'll buy out your licenses.
We'll buy out your hardware.
We'll make it easy on you butwe sell networking.
We don't sell Wi-Fi.
We don't sell routing.
We don't sell switching.
We don't sell security.
We don't sell DNS security.
We don't sell SD-WAN.
We don't sell VPN.
We don't sell ISP management.
(38:16):
We sell networking.
Speaker 2 (38:18):
That's great.
I mean, I actually think that'sa great answer because I think
if you, as the vendor, if youcan stand firmly on what your
product is and what it isn't andyou stick to that, then I think
that goes a long way.
So I actually like that answera lot.
Good, a long way.
(38:38):
So, um, I actually like thatanswer a lot.
Good.
Um, I guess kind of compoundedwith that like what, what do you
tell customers that are kind ofon the fence like hey, I don't
know if I want to completely ripout my stack and put in
something brand?
new right, you know, I mean yeah, I mean, if we use the apple
example, there are there'spluses, there's minuses to kind
of a I won't say a closedecosystem, but you kind of know
where.
I won't say a closed ecosystem,but you kind of know where I'm
going right A full stack that'sdriven by one vendor, right?
(39:01):
So what is your kind of mainselling point that you would
articulate to prospects, newcustomers in that camp?
Speaker 3 (39:08):
So I think it's three things.
It's P P, p, product process,pricing.
First, on the product sideperformance, reliability,
security and integrated products.
I think the delight of beingable to go to our DNS security
and saying you know, andy, can'twatch any more TikTok that
(39:31):
being a single line, it could becrushed.
That being a single line, itcould be crushed.
That being a single thing to doand not having to go buy a
separate DNS product andfiguring out how to integrate it
into the firewall.
And what's going on with theother DNS?
How's it going to propagateproduct at the end of the day?
How well we perform.
If you look at even VPNperformance, we blow away
(39:54):
everybody in the industry.
Straight performance we blowaway everybody in the industry.
Straight performance you knoweverybody talks about is nobody
does benchmarks anymore Because,guess what?
Everybody's asymptotically gonetowards mediocre stuff.
As meter gets more out there,you'll see us actually do
benchmarks.
We believe what we're buildingand we know what we have.
(40:14):
That's much better, so muchbetter reliability, security,
performance, integration andbeautiful products.
That's the first one.
It's on the product side.
Second is on the process thatis delivered because of the
products.
Like I mentioned, right when acustomer signs up.
We have our connect productthat enables us to help them
(40:34):
procure ISPs.
Then we have an entirelyonboarding product where
networking and IT teams can inline, answer a bunch of
questions on how they want thenetwork to be.
All of it, every single pieceof our hardware, down to a port
on a switch, is virtualized inthe backend.
This has become like an overusedterm now digital twin kind of
(40:57):
thing, but in practice I'll tellyou what it means for us, which
is every configuration of anetwork can be actually
configured in the backend, inthe cloud, before it's applied
down to the hardware.
There's no way of bricking thisthing, and so going from ISP
onboarding, configuration,integration and deployment of
(41:19):
networks for a lot of customersprocess is what's attractive.
And the last one is pricing.
We generally in networking it'sbeen huge CapEx and huge OpEx
with licensing and renewals andall these things to purchase
networking gear.
We are zero CapEx and if you dothe TCO out over the course of
(41:45):
three, five, seven years, it isremarkably in the favor of meter
customers.
So depending on who you are,how you're set up and where you
are in the life cycle, it couldbe some combination of product
process pricing.
Speaker 1 (42:02):
Yeah, got another hard question.
Sure it's not really hard andI'll set us up with you know.
So again, I think it'scompelling.
Even the ISP marketplace, likethe first stop that you hit in
this experience I've never seenanything like that.
We at places I've worked, wehad entire teams of analysts
(42:24):
that did all the telco stuff forus because we were busy in the
iron.
So the fact that you can justgo to a screen I thought you
said that it's open to you youdon't have to be a customer to
even you don't have to be acustomer to even you don't have
to be a customer.
Speaker 3 (42:37):
It's what we're providing to the industry.
Please go to metercom slashconnect.
It's called Meter Connect.
Use it, give us feedback.
We'd love to hear from you.
Anybody can use it, send it toyour friends, send it to your
aunts, send it to your kids.
Speaker 1 (42:51):
And there's pricing in there too.
Right Like I can go in and Ineed an MPLS.
Speaker 3 (42:54):
Go from saying here's what I want from 5G coax,
shared fiber, dedicated fiber,starlink.
Here's the pricing.
Just give me a contract thatnormally would take weeks what I
just mentioned Absolutely To beable to do.
(43:18):
So all of that is in there.
Connect is super loved product.
Speaker 1 (43:20):
Yeah, andy, I think your kids would love it too, for
sure, so I can order it rightfrom there, right?
And then does that kick off.
That's right, I mean, I'll getan install date, or right.
Speaker 3 (43:31):
That's right.
Speaker 1 (43:32):
It just happens, that's pretty.
That's right Again.
For anyone who's had to try toorder a circuit, this is very
simplified process.
It's it's a.
It's a logistical nightmare theold school way.
Um, again, hire teams of peopleto do all that and and manage
the billing and forget it rightyeah so.
Speaker 3 (43:52):
so that's for sure.
For one location where youdescribed as soon as you have
multiple locations that havecome up in different times
renewal management, contractmanagement, that stuff is harder
Then.
What you're also talking aboutis I now have 500 locations with
a primary and backup.
That's a thousand ISPconnections I'm managing and I
(44:15):
have to pay.
So Connect has that buying,contracting, consolidated
billing, so you can get a singleline item.
We can pay and make that easierfor you, and because we buy so
many connections, in majority ofthe cases we get you better
pricing than you can ever get.
Speaker 1 (44:35):
That's nice and the billing's integrated right,
Because I can't tell you howmany outages I work because we
forget to pay the bill.
Right, I mean, it's a thing.
Speaker 3 (44:43):
I hear that all the time.
Speaker 2 (44:44):
Yeah, okay.
Speaker 3 (44:46):
So what does?
Speaker 2 (44:46):
support.
Look like in that scenario?
Like, if you know there's acircuit outage from something
that you've procured throughConnect, who do you call?
From something that you'veprocured through Connect, who do
you call?
Speaker 3 (44:54):
So it's two separate models.
If you have our entire network,there's one support.
If you're just buying the ISPusing the marketplace, it's a
different one.
If you have our full network,we have host monitoring built in
.
We have ISP monitoring built in, because it's art of networking
.
I'll give you a preview onwhat's coming.
(45:15):
We will automatically track howyour ISPs are doing against the
SLAs you signed as well.
That's coming.
So if you have the entirenetworking stack from us, we can
keep track of the whole thing,and then we're the ones doing
through our APIs and contacts.
We know when it goes downalerting that comes in.
You get the full service.
(45:35):
But if you just use themarketplace to buy, that's
whoever is on your team, howeveryou want it.
We just made that transactioneasier for you, but we're not
responsible for that circuit.
Speaker 1 (45:48):
Are you monitoring utilization?
Oh yeah, If I'm gettingsaturated and I should upgrade
and all that stuff, Becauseagain that's a whole other thing
, right?
Speaker 3 (45:56):
Oh yeah, we proactively let our customers
know that you're saturating it.
Here's what's happening.
This recently happened with uswith a robotics company Because
of how they were gettingthrottled.
I think we ended up finding abug in the I won't say which
carrier, but in the carrieritself.
Then we were able to do that,but we proactively, a lot of
(46:17):
times let our customers know.
I think the times we'reactually most proud of this
happened recently, when we tellour customers they're paying for
a much bigger circuit than theyneed and say here's all the
data from the last six months ora year You're actually spending
$2,500 on that primary circuit.
You don't need that two gigcircuit, you're okay with one
(46:38):
gig.
Save a thousand dollars a month.
Those are the ones we'reactually the most proud of.
Yes, we tell them when theyshould spend their most money,
but I think it's really nice tobe able to go to a customer and
say please spend less.
Speaker 1 (46:49):
The hard question I never got to.
I was trying to set it up and Idid a poor job doing it, but
this is obviously a lot ofbenefit for businesses, right?
So this is a show consistingmostly of network engineering
folks.
So and you and I had talkedabout this before we started
recording I seem to have how canI say this and not sound like a
(47:16):
big wimp but I becomethreatened by things that
disrupt the model.
Let's say that.
So you know, automation, I felt, was a threat to my job, right,
cloud, I felt, was a threat tomy job.
Now, as it turns out, as I'velearned over time, that stuff's
additive, right, that it's notthat it's taking away my job.
(47:36):
It's that these things likeautomation and cloud can
actually make my job better oreasier or more efficient, or I
can offload things for mycompany in the cloud, but it
doesn't mean they're firingother networking people.
So, to bring this full circle,meter is doing my job, I think,
and I am threatened by it.
So what do we say to networkengineers who look at your full
(48:00):
integrated, you design, youdeploy and you manage the
network, which is, I think, whatmy job is?
What do we say to those folkswho might be like, well, this is
great, but this is a networkengineering show and they're
taking our jobs?
Speaker 3 (48:16):
Yeah, and I think it's a valid question.
You know, andy, you'resomewhere in mid thirties.
You know you've been doingnetworking for the last 15 years
or something.
But I think the reality is thisyou know, I studied networking
about 20 years ago.
Now, when I go back to theuniversity where I studied, they
(48:40):
don't even offer the networkingdegree anymore.
That's just a fact.
As a network engineer, we haveto come to terms with this fact.
I recruit from some of the bestuniversities all the time
Stanford, mit, whatever.
These kids don't even takenetworking classes because
they're not even offered thatmuch anymore.
They get one week of networkingin a distributed systems class.
(49:01):
So I think the first thing toactually realize is the inflow
of talent just isn't there.
That's one.
Two, it's been the same numberof network engineers for a while
.
There's actually a massiveshortage.
If you look at companies likeGoogle that have Google
University and other things,they're trying to train others
(49:24):
into networking because the flowisn't just there.
I'm happy to send you this link, but you should look up Bureau
of Labor Statistics data on iton how many software engineers
have grown network engineerswhich don't have the inflow
that's coming in.
So it's roughly been the samenumber of network engineers for
a while.
But coupled with that, there'stwo things that are happening.
(49:45):
One is the number of devicesand applications on a network
have grown dramatically.
Network have grown dramatically.
Two it used to be that therewas one IT or networking person
for every 75 to 100 people in acompany.
Today, the average is about 250.
So you are in a situation wherethere's no new inflow of talent
(50:08):
and the existing people askedto do way more than they
currently are.
And the network has become moreimportant than it's ever been
because everything in the worldis packets.
Now, as we all get old andactually studied networking, I
don't know who's coming behindus, but even the folks that are
today doing it.
(50:29):
If you ask them hey, do youhave time for that project you
want to get to, or anything, orhow are your KPIs?
How are you dealing with it?
When was the last time youmonitored the network?
When was the last time you didupgrades?
It's a lot to be able to dowith no new talent coming in,
more expectations and morenetwork load than ever before.
Speaker 1 (50:50):
It's untenable, right Like we seem to be heading
toward catastrophe.
Speaker 3 (50:55):
So that's actually another huge premise of why we
started Meter is, if you're inthat position, we did not feel
like there was a company thatwere building hardware and
software for me as a networkengineer, for me to be able to
do my job well, which is servemy colleagues so that the
overall company can do betterNetworking is this Archimedes
(51:15):
lever for the rest of thebusiness?
One of the things I really careabout is making networking cool
again, like truly, but I dothink companies like us.
We hope to be augmentingnetwork engineers so that they
can do much better things fortheir customers, but the reality
(51:36):
also is there's not enough ofus.
Networks are falling apart.
There's end of life, end ofsupport networks out there, with
some of the largest retailerswith credit card data flowing
through that who knows when wasthe last time they did anything
with it.
So I hope what we actually dois build such great hardware and
(51:56):
incredible software.
That's why even something likeCommand came in.
Is that we want to buildsomething so good that all the
folks that are young, like you,andy, that grew up doing CLI,
would love to use Command.
Speaker 1 (52:09):
Just keep calling me young man.
No, that makes a lot of senseand it rings true to a lot of
the things I've been hearing forat least the past year or two.
We don't have new talent comingin.
I can't believe.
Yet I keep hearing that thatcomputer science courses have
(52:30):
basically networking from thecurriculum.
You said something earlier onthat rang true for me when I was
a cable guy at an ISP and I sawthe statistics, the you know.
You know today we have fivewire.
You know five computing devicesin a home and you know, in two
years we'll have 10.
And there was kind of thisMoore's law type math of the
(52:52):
explosion of devices just inresidential areas.
And that's when I thought, wow,networking will grow for quite
some time and I will definitelyknow this is a good career to go
into.
So it's such a strange thing tohave that you said it before
that explosion of end users andeverything is on the network.
(53:13):
I mean we have thermometers andfish tanks now that are IoT.
There's just so many devicesand the fact that we don't have
the talent coming into theindustry to support all this and
it's just going to keep growing.
So I'm comforted that a companylike yourself is addressing that
?
Because it sounds to me like ifI am a network engineer working
(53:36):
at a company using meter, I cando a lot more with less people
like myself.
I mean, we have less of us.
So I guess we need tooling likemeters to be able to manage a
huge infrastructure, get moreinsights, that more kind of
integrated solution.
So isn't it funny I'mthreatened by meter and yet
(53:58):
you're trying to help me.
So I'm glad we're having theseconversations out loud, because
when I see this I'm like, ohgreat, as a service, they're
trying to make me go away, butthere's less of me every day,
every month, every year.
Like we need help.
Speaker 3 (54:11):
Yeah, I mean, we think of network engineers as
farmers and we're just shovelsImagine doing farming without
shoveling or a tractor or a plowand we felt like you know,
there were some legacy vendorsthat build a shovel, a plow but
it kind of stopped there.
We need tractors for networkengineers and it needs to be
(54:34):
really great products.
Again, the simplest way for youto know if people have cared
about networking the decade ortwo is if you pick up a
networking product that might behardware or software, do you
actually find it beautiful?
And I'm sorry there's not asingle beautiful networking
product out there until we camealong.
That tells you how much peopleactually care and we want to
(54:59):
build that.
That lever is important.
That productivity multiplier isimportant.
Speaker 1 (55:03):
And this has been very informative for me.
We got some great interactionsin the chat.
I think we found you a DAScustomer, so that is awesome.
Any parting thoughts?
I think we covered.
You know all the questions Ihad and then some.
I think what Meter has iscompelling.
(55:24):
It solves a lot of problems fora lot of different use cases.
Anything before we close thisout.
Speaker 3 (55:30):
Yeah, I think three things that I'll mention.
One we really care aboutnetworking.
There should be great hardwareand software.
Hardware should not becommoditized.
Software should not be ugly.
Two, we do everything from ISP,routing, switching, wireless
DNS security, sd-wan, vpn, allintegrated.
(55:53):
That's the way it's supposed tobe.
And three, I love networking.
So if anybody wants to reachout, my email is anil at
metercom M-E-T-E-Rcom.
I'll respond pretty fast.
Speaker 1 (56:07):
Awesome.
Thank you so much, anil.
This has been a great session.
Chris, thanks so much forjoining.
If you want to check out Meter,you have Anil's email address,
which is pretty cool.
I don't have many emailaddresses of CEOs.
You're going right in mycontacts.
I'm going to bug the heck outof you.
You can also find them atmetercom.
They have that awesome freeresource which I'm trying to
(56:29):
find here.
It's metercom.
Speaker 3 (56:31):
Where is it?
Slash connect.
Speaker 1 (56:33):
Slash connect, you can check out their marketplace
of all kinds of ISP connectivity.
I love it.
And to check out everything Artof Network Engineering, you can
check out our link tree atlinktreecom.
Forward slash, art of NetEng.
Great conversation.
Thanks so much, anil Chris, andwe'll see you next time on the
Art of Network Engineeringpodcast.
Engineering Podcast.
(57:00):
Hey everyone, this is Andy.
If you like what you heardtoday, then please subscribe to
our podcast and your favoritepodcatcher.
Click that bell icon to getnotified of all of our future
episodes.
Also, follow us on Twitter andInstagram.
We are at Art of NetEng, that'sArt of N-E-T-E-N-G.
You can also find us on the webat artofnetworkengineeringcom,
(57:21):
where we post all of our shownotes, blog articles and general
networking nerdery.
You can also see our prettyfaces on our YouTube channel
named the Art of NetworkEngineering.
Thanks for listening.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Intentionally Disturbing
Join me on this podcast as I navigate the murky waters of human behavior, current events, and personal anecdotes through in-depth interviews with incredible people—all served with a generous helping of sarcasm and satire. After years as a forensic and clinical psychologist, I offer a unique interview style and a low tolerance for bullshit, quickly steering conversations toward depth and darkness. I honor the seriousness while also appreciating wit. I’m your guide through the twisted labyrinth of the human psyche, armed with dark humor and biting wit.