August 2, 2025 • 35 mins
In this episode, David Yakobovitch from Data Power Ventures joins Joel Palathinkal to discuss the evolution of big data players going public and his career journey, including his current role at SingleStore. David shares insights into investment strategies and industry experience in data startups, highlighting the data science investment landscape and exciting segments. The conversation explores the future of quantum computing and offers advice for aspiring investors. David emphasizes the importance of partnering with accelerators, LP relationships, and trends in DevOps and cloud strategies. The episode also covers advances in data storage, biopharmaceutical research, breaking into venture capital, and building a strong community, concluding with mentorship insights.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Lot of these big data players who have scaledare now going public, but that doesn't mean
(00:05):
that the war has been won.
In fact, it's only the early days because mythesis is just how we looked at the twenty tens
where every company is a software or technologycompany.
Now it's the world where every company is adata company.
Welcome to The Investor, a podcast where I,Joel Palafinkel, your host, dives deep into the
(00:30):
minds of the world's most influentialinstitutional investors.
In each episode, we sit down with an investorto hear about their journeys and how global
markets are driving capital allocation.
So join us on this journey as we explore theseinsights.
PhD.
That works for me.
(00:53):
Less syllables.
Cool, I think we are live now.
So hey, David, really excited to have you onthe show.
David and I have just become friends over thelast probably six, seven months and excited to
have him in our fund accelerator as well.
But really just, it's interesting to see somany different funds now from my experience
(01:17):
break out into specific niches.
We've had psychedelics funds, we've hadhealthcare funds, climate change funds.
You're the first fund that I've met that has astrict focus on AI, ML and data science.
And I would say, this is something that youcould probably educate us on, really just the
trends and the exits.
Because I feel like some of these large datascience companies are the ones that are really
(01:40):
getting acquired by massive, probablyinstitutions and then larger conglomerates like
Amazon.
So I'm excited to kind of learn a little bitabout that as far as the trends and exits and
then just the space.
And I think another big thing that would behelpful even for me and the people here is to
just unpack data science.
So there's AI, ML, there's explainable AI,that's another hot topic.
(02:06):
But before we nerd out on all that stuff, whydon't we talk about you, David?
I know you're from Florida like myself, solet's go back to there.
Let's go back to where you grew up, where youstudied and how that helped you pivot into VC.
Yeah.
Thanks, Joel.
Thanks so much for having me on the show andproud Floridian here.
(02:26):
I like to always say, go Gators, the Universityof Florida alumni.
And for me, it does go back to the collegedays.
In fact, I was studying actuarial science,finance and information systems, wanted to go
hardcore into data.
And I discovered in my first internship where Iworked for Aflac doing loss experience
(02:48):
monitoring.
And and that means to teach me like I'm five,that that means actually predicting what the
price should be for your insurance, whetherthat's car insurance or health insurance.
And I discovered at that time that the cloudwas emerging and that no longer did working
with algorithms have to be in Excelspreadsheets, but could be in large distributed
(03:13):
systems and with programming languages likeSQL, Python, and R, which tapped me into
wanting to evolve into more of the data field.
And so I've spent now over the past decade,working for both the big Fortune five hundreds
as well as start ups in all things data.
I led and scaled data analytics and dashboardsand products, beyond Aflac to, Deutsche Bank,
(03:40):
Citigroup, and ADP, and that was all inFlorida.
I worked for some great organizations there andand then, of course, decided to take my startup
plunge, into the startup ecosystem when I movedto New York City about eight years ago.
And tell us about moving to New York because Ican probably really relate.
(04:01):
It's a different ecosystem, it's a differentenvironment as a whole.
Mean, people in general might call us too laidback or just say that we're a little more less
hyperactive as the typical New Yorkers.
So, tell me about just your thoughts on NewYork and what made you think about New York and
(04:26):
then, you know, just what you experienced whenyou first moved there from Miami.
Sure.
So I did the reverse migration.
You see many people, especially in the pandemicor or later in their life where they moved to
Florida, and I decided that moved to New York.
And for me, was fascinating because all thecompanies I worked for, Deutsche Bank,
Citigroup, ADP, all have New Jersey and NewYork offices.
(04:49):
So over my first part of my career, was goingtwice a year up to the offices and saying, what
is this ecosystem?
What am I missing out on?
In fact, I also considered moving to SiliconValley, and I spent a few months living in the
Bay Area saying, I'm thinking about SF or NewYork.
And I landed on New York for a few primereasons.
(05:09):
The first is the diversity of the ecosystem.
You have every industry imaginable in New Yorkfrom business, technology, fashion, and
finance, and that creates a wealth of newstartup ecosystem, innovation, and ideas.
And you also have a wealth of individualsworking in different careers.
You have the sales development reps and accountexecutives.
(05:31):
You have the engineers, the CEOs, and thefinance team members.
In Silicon Valley, I only saw engineers.
And for me, it was important to align close tothe business, which led me to make it to New
York.
And I think one of the best things about beingin New York is the fast moving pace of the
lifestyle.
I've traveled to different cities around theworld like Singapore, Taipei, Paris, London,
(05:55):
and nowhere is it as fast paced in New York.
So today when I work with founders andinvestors, I'd say there's nothing like having
a New Yorker on your cap table.
Yeah.
No.
I totally agree.
And then tell me about the the startup you'reworking for and what your role was there.
So today I work I I continued going into thestartup ecosystem once I moved to New York.
(06:17):
Mhmm.
And two of the startups I helped scale wereGeneral Assembly and Galvanize, which were
leaders in the EdTech ecosystem.
And I was scaling their enterprise data scienceteams, which meant I was supporting Fortune 500
clients like Charles Schwab, Bloomberg,Refinitiv, Invesco, USAA.
And after both of those companies wereacquired, I joined a little over a year and a
(06:39):
half ago, SingleStore.
And SingleStore is a database startup.
It was Y Combinator 2011.
We're backed by Google Ventures, Insight,Cosla, Dell, and HPE.
And today, eleven years into its journey, inthe last eighteen months, SingleStore raised a
series e and series f finance and is continuingto scale on its pathway to being beyond a
(07:03):
unicorn.
We're on that that territory, but primarilyhaving the passion and purpose to creating an
industry where any company can be a datacompany.
Where companies no longer have to think aboutwhere their data is warehoused or what data
lake the data is being managed on, but thateveryone can build for real time applications,
(07:24):
for data intensive applications, and for fastanalytics.
And so my role being there has been scaling ourtechnical enablement that's been with clients,
partners and employees, as well as looking forstrategic initiatives for partnerships.
You make a, you know, and you're a perfectexample.
I mean, there's a lot of people that havebecome an investor with the industry experience
(07:47):
that they've had.
So do you think that's a good starting pointfor a lot of people if they work in healthcare,
if they're working like healthcare IT to maybejoin like a digital health fund?
Or do you think it's also good for people tojust kind of jump in to other industries as
well?
And it sounds like for you, I mean, that's kindof been your niche, right?
You really built a lot of great enterprise andindustry experience in data.
So it just kind of makes it a natural fit tojust, you've already understood all the
(08:09):
problems in the business models to to just jumpin and directly be on the investor side.
Yeah.
I really appreciate what you shared earlier,Joel, which is about niching down into a
vertical.
And for me, it's just well, my whole career hasbeen around data ML AI.
So it made sense.
Take what I know and apply that to a diligenceand to mentorship and advisory for startups,
(08:34):
which led into investments.
So I do think, if you're someone who's emergingas a fund manager, if you're getting into the
angel investing industry as well, start withwhat you know.
It is a vast industry.
When we started, our thesis was about beingdata driven, but we realized very quickly that
was too general, and so then we narrow downinto data intensive apps, real time insight,
(08:59):
data developer tools by focusing on data power,those infrastructure layers, those insight
layers, and those prediction layers of theeconomy.
I think that's really helpful when you're a newinvestor because otherwise, you're seeing so
many opportunities, and it's tough for you tobuild your mental models.
You wanna build the best mental models so youcan decide how do you make investments.
(09:22):
And I think that's based on either theknowledge you know or the team you surround
yourself with.
I've also heard, you know, I just wanna make acomment to your earlier point.
So with the niche funds, I've I've also seenthat be really attractive to LPs because LPs do
have different interests.
They're interested in Web3, they're interestedin quantum computing, they're interested in
(09:43):
psychedelics, but they don't have the time tosource these deals and they don't have the
network and they may know, they may not havethe technical chops.
A lot of them do because they're pastentrepreneurs, but a lot of them just may not
have the technical chops to due diligence.
And one thing that you and I talked about,which helped you, I think, form a really
interesting partnership with someone was yourability to kind of do technical diligence with
(10:05):
someone else.
So I think that's an interesting point.
LPs want exposure to an exciting sector that isreally hot, but they just don't have the
bandwidth to go out and source all those dealsindividually.
And my favorite analogy is, you can either buyTesla ad hoc or you could just buy it index
with like a bunch of the FANG companies in acertain theme or something like that.
(10:29):
So I just wanted to make that comment, but youstarted doing this.
I want you to go a little deeper.
Can you just unpack the whole universe for usfor just data science and how the investment
activity is broken out?
So we talked a couple of things we talkedabout, right?
Explainable AI, AIML.
Can you just kind of give us a high leveloverview of just the whole landscape and then
(10:50):
and then where you play?
You've talked about like the data power.
So maybe just kind of elaborate a little morefor our education.
So when we think about software and data, thebig trend to look at is the emergence of the
software engineering industry.
And so software engineering had its rise tofame in the early two thousand.
(11:11):
I would say that even went to, like, 2010 and2015 with software automation and all the tools
so that every engineer has a toolkit where theycan build the best world class systems.
What we're seeing with data is that has notexisted until now.
The twenty twenties are known as the decade ofdata.
This is where the modern data stack's beingbuilt.
(11:33):
New developer tools, new infrastructure, newsystems so that data scientists, ML engineers,
and AI specialists alike can have their owntoolkit to also build and scale enterprise
grade data systems.
And to do that, it doesn't only require thetools, it requires the systems and having them
talk to each other.
(11:53):
So we've also been seeing as a result of that,some of the early players like the single
stores who've been around since 2011 to 2015focused on data are now going through m and a,
getting acquired, and or IPO ing.
We've seen companies like Datadog go on to thepublic markets, MongoDB, Snowflake.
(12:14):
A lot of these big data players who have scaledare now going public, but that doesn't mean
that the war has been won.
In fact, it's only the early days because mythesis is just how we looked at the 20 tens
where every company is a software or technologycompany.
Now it's the world where every company is adata company.
(12:38):
That's exciting.
And what are some of the the segments thatyou're most excited about kinda unpacking that
with the with the data, focus, I guess?
So I look at the data industry as four layers.
The first layer is data collection.
And data collection has typically been a verymanual process with labeling and aggregating
(13:01):
the datasets and then storing them.
So that's beginning to automate.
That's a very exciting space.
It's growing a lot.
We've seen Labelbox, Scale AI, and otherplayers become unicorns in that space.
We even see newer companies like Roboflow,which are helping automate some of the labeling
for computer vision and NLP systems.
(13:21):
Then we have layer two, and layer two is reallyabout the real time insights.
It's having tools that are dashboards orspecific for different industries, whether it's
health care, biomedical, or looking at securityin in the cyber industry and having these tools
so a company can say, I can buy this tool andhave data layer integrate with the APIs and get
(13:46):
real time insights without having an army ofdata scientists and data analysts to generate
these insights.
So that's layer two, which we're seeing a lotof platforms emerge and scale in that space.
Layer three is where we start seeing thepredictive insights.
That's the traditional machine learning that'sbeen told of, if I apply for a loan today, do I
(14:07):
get approved or do I get rejected?
And looking at those insights, in real time isbeneficial for industries to not have to have a
delay of taking a few days for results, butinstead having instantaneous data intensive
apps.
And so we're seeing that today in underwriting.
We're seeing that in credit reporting.
We're seeing that in these new, ML powered appsthat support, you know, your need to be more
(14:34):
successful at enterprises.
And then the fourth layer is about AI.
So most companies today talk about we're AIfirst, but if you don't get the other three
layers right, then the fourth layer won't getthere just yet.
And so the fourth layer of being AI poweredmeans you're using this modern AI stack, which
(14:54):
could be like OpenAI's GPT three to generatenatural language processing text or create
generative images or create systems that selflearn and heal on themselves.
The challenge when companies talk about beingAI powered is it's the whole ecosystem of data,
ML, and AI, not AI alone.
(15:15):
And so I'm really excited about all four ofthose layers, and we looked at all of them with
our investment thesis and mathematics.
It's really exciting, and that that was reallygood learning for me.
Where do you think quantum will play in all ofthis?
I guess will quantum have a because from mylearnings, I'd love for you to clarify or help
me with this, some of the bigger applicationswith quantum are like material sciences,
(15:41):
infrastructure.
There's also some chemical applications aswell.
So, I guess there's applications for all ofthese that you mentioned, right?
The collection, the automation, the predictiveand AI power stacks, but how will quantum
evolve into kind of this whole ecosystem?
(16:02):
I'm very excited for quantum and I think it isstill the early days of quantum.
I do have several good friends who work in thatspace.
So we routinely chat about the trends andchanges that we're seeing in the quantum
industry.
Just this month, I've looked at this reallyexciting quantum startup that's using quantum
entanglement to replace global GPS withsatellites to get better real time data for
(16:28):
clocks and movement of trains around the world.
And when you you hear this story for the firsttime, most people say, is this really a
problem?
But the truth is there is.
There was cases where thirteen microsecondswere off between some of these, travels, which
caused flight airline, disruptions andinterruptions.
You even have trading systems where there'sthirteen microseconds that can mean the
(16:50):
difference between a profitable or a lossdriven month.
So I think there's a lot of opportunity inquantum, and we've seen that with the cubits.
We've seen our systems scale to beyond a 100cubits.
I I think we need to get to that thousand milemarker to start seeing some benefits just like
we saw in the AI space with the scale of GPUsand now with ASICs in in the blockchain space.
(17:15):
So Quantum will have its day, and and therehave been some unicorn IPOs there as well.
I just think it's still the early days, and weneed to find commercially viable business cases
to work with Quantum.
Yeah, no, it's really exciting.
What are some pieces of advice that you wouldhave for people if they're interested in being
data science investors for sourcing andscreening?
(17:37):
Are there a lot of these, are they still in theacademic ecosystem?
Because I know a lot of the quantum deals andthe technology is still in like tech
transformation.
Some of the events that I went to in the pasthad a lot of university presence kind of doing
a lot of research, but is that the similarspace to kind of meet a lot of these data
startups?
(17:58):
Are there great accelerators that areincubating these startups and technologies?
I guess what are the best avenues to kind oflearn about these opportunities?
Yeah.
So I do partner with a lot of accelerators andsome of my favorite ones, of course, are
Techstars being in New York City.
I've seen a lot of companies emerge.
So they're building developer tools around dataand ML in the Techstars ecosystem.
(18:23):
Of course, there's a lot of ways that you candiscover the market because there are so many
data startups today.
In fact, my mission is over the next, ten yearsto accelerate a thousand data powered startups.
So I think we're gonna continue to see morefounders and more, executives and engineers
alike move into the data ecosystem.
(18:44):
But when all things are considered, if you'relooking to invest, I I think it goes back to
your point earlier, Joel, that there's a coupleways.
Right?
You can go deal by deal or you could do thatindex portfolio.
And that's what we see with a fund like DataPower Ventures Mhmm.
That we've had LPs come in really for a couplereasons.
First, they're either the engineers or businessexecutives who build or scale technology, and
(19:05):
they understand that data is the next wave thatwe're seeing in industrial revolution.
Or secondly, they want that exposure.
They want that education.
They're a finance executive who doesn't knowmuch about data ML AI, but they also believe
that, hey.
This is something that I don't wanna miss outon.
I want the education information rights and tohave also some upside in my portfolio.
(19:27):
Yeah.
No.
I totally agree.
I think those are all, good reasons to becomeLPs.
And it's also just, it's great too if you canbuild a relationship with LPs and have these
focus groups and kind of keep them involvedbecause a lot of them have that industry
experience as well.
Some of them are really well known datascientists.
(19:49):
So, they may actually help you de risk some ofthe investments as well.
So, I think if you keep them engaged and kindof seek feedback, that also helps to build that
long term relationship as well.
That's just kind of what I've seen.
And I can probably assume you agree too.
Community is everything.
Absolutely.
(20:09):
Yeah.
David's got a really great, I wanted to plugthis too.
You've got a really amazing community oftechnologists as well.
You have a tech dinner that you host every oncein a while.
Is there a website where we can go to sign upfor that?
Or is that just more off the record kind ofconnecting with you?
Yeah.
So there's a few community driven events that Ido, which I can share.
(20:32):
You know, number one, with, the podcast, I dohave an AI data science podcast been running
for over three years called Humane with AI inthe middle.
So that's h u m a I n.
And we have Humain podcast where you can listento great episodes with founders from Preseed to
IPO Ventures.
Data Power is at datapower.vc.
(20:52):
So you can, of course, learn more about ourportfolio and and how we're investing and and
everything around there.
And the Tech Dinners, this is an invite onlycurated community around investors, founders,
developers in New York City.
But to get involved with the Tech Dinnerseries, you can find me on LinkedIn, and you'll
(21:13):
see my most recent saved item on my profile hasan Airtable link.
So I'm making it really hard for you to get to.
But if you're interested and excited, you cancheck it out.
And we we run these events in New York City.
We've actually brought multiple portfoliocompanies in person with LPs.
They've directly invested after meeting thefounders.
(21:35):
We've brought in near introductions who've beenhired by the portfolio companies.
So they're very strategic community drivenevents.
Yeah, I mean, I've been seeing a lot of peopleuse NFTs now to kind of have utility in the
community and also get access to events aswell.
So I wouldn't be surprised if you had a datapower NFT launching at some point in the
(21:55):
future.
Let me think, what are some other hot trends?
So I think, one thing that I've seen too, andthis probably falls into the tools, I've seen
some platforms that help to just provide costsavings.
So like, I think Cloud Admin is one of themwhere they help to kind of help you manage your
costs.
So do you think that's probably a hugeapplication and very, very high value to data
(22:19):
scientists, but do you think that's a hotindustry or there are a lot of people doing
that now?
So we've seen a few a few startups in thisDevOps automation space in the last couple of
years.
And and I agree that this is very important.
Even at single store, we face these costs everysingle day because we're a cloud first company
and our bill is, at minimum, into the hundredsof thousands of dollars a month on cloud costs.
(22:44):
And just by using some automations that we setup recently, we're now saving, at least a
$100,000 a month.
This is no joke.
Right?
And this is just by an engineer writingscripts.
So there are startups that also are, focused onthis business where they can plug in to your
cloud, AWS, GCP, Azure, and then helpunderstand think about these systems turning
(23:06):
off, turning on, or using more reserved ordedicated instances, and you can see those
changes.
I think that's very important, because one ofthe underlying factors that we've seen as a
result of the pandemic is that every company isdigital first.
You no longer have a choice.
And even in the hybrid world, that means cloudis here to stay, and a cloud first strategy is
(23:29):
required.
Yeah.
No.
I totally agree.
Where do you see data?
So we've seen a lot of these companies do likeUI automation.
So what's kind of like super deep tech for you?
Like, do you feel I'm trying to think of likesuper sci fi stuff with data.
So I think when I can think about it, I canthink about like just real time movements of
(23:50):
your body and like maybe predicting.
What I think would be really interesting is ifyou could, and they're probably already doing
this, but based on your family history andbased on kind of your eating habits in real
time, like imagine if they could just detectyour lifespan.
If you still eat pizza and fried chicken likefour to five times a week, they could probably
(24:14):
already predict some outcomes because there'sso much data that's been collected.
There's probably like some parts of it too tiedto like your ethnicity and your cultural
background too.
But I feel like that would be a really hotspace for just like real time data And then
predicting kind of what your future be ifyou're continuing to do what you're doing.
(24:34):
There could be results.
There was a company that I that I met three orfour years ago.
And they were trying to do that.
I don't know if they got the traction, but theyclaim to be able to detect within high
accuracy, like when you would get a heartattack based on your activity.
But I don't if you've been seeing anything likethat in the health care space, but that's got
to be a huge insurance play as well with youcoming from like Aflac as well, maybe.
(25:00):
There there's so much opportunity in the deeptech sci fi space that is really fascinating.
And, companies we've looked at, that we haven'tparticipated in, but we've we've explored, one
includes a a satellite company that's producingsatellites at one tenth the cost.
But what's great is the satellites are trackingdata about where objects are in space to then
(25:21):
provide the insights for, space trash removaland clean up.
So that's a really, deep tech play.
Oh, wow.
Another one is a a sensor company that installsin warehouses these special sensors to see
where any movement occurs anytime based onvibrations and waves occurring in those
(25:41):
warehouses.
And this is a great company because, you know,Boeing and Maersk and others have already
committed MOUs to them.
So we're really looking at that kind of deeptech to see all these motions.
So I think those are a couple really uniqueones.
One that's more down to earth on, literallyspeaking on the health side, that we're
finishing our investment in, is is focusedaround accelerating research and development
(26:06):
for biopharmaceutical companies, and thatcompany is called Applied XL.
And so Applied XL, is led by, Francesco, who'sthe former head of r and d at the Wall Street
Journal.
Really exciting startup that we're finishing toround out with Data Power Ventures.
And and that company actually, is is alreadybeing led by some really exciting, investors
(26:28):
all in New York City.
So we do love to support New York founders.
We've actually have five of our portfoliocompanies are based out of New York.
So there's, I think a lot of movement we'reseeing today in the entire data ecosystem.
Yeah.
And then I think one thing that I've seen inthe last probably five or six years is just
kinda advances in how you're housing the dataand storing them and categorizing them.
(26:49):
So maybe you can share a little bit of insighton like just some of the enterprise workflows.
Like now they have the concept of a data lakewhere you're kind of taking the data
categorizing in this lake where you can kind ofuse it whenever you need.
But can you educate us a little more on kind ofsome terminology that like if you work at a big
data company, some things to think about, ifyou're like a data scientist on the enterprise
(27:12):
side, people may not know about it because theydon't work at these companies.
So but I think you might have a little moreinsight into just kind of the industry because
there's just a lot of internal workflows thatpeople need to think about.
At sales store so we we call it actually threetiered storage.
So this is our our name of the technology,which is similar to data lake.
(27:34):
It means that, one, you actually have the datastructures that you house the data in.
For us, that's row stores and column stores.
For other companies, those are data frames.
Those are CSV spreadsheets.
Those are SQL tables.
Then below that, we actually have the storageon your hardware.
So that could be, in this case for us, theseare the compute units, which are Linux AMI
(27:57):
machines, and they have their own, databasethat's sharded or distributed, into storage on
these machines.
And then where the data lake appears, we see itwith Databricks, where they have their
universal data lake and their Delta Lakesystem.
And then for single store, it's where we havewhat's called bottomless or unlimited storage.
So pretty much, it's like you have this oceanof data, and you can dip into it with a fishing
(28:22):
net and grab the data that you need when youneed it without searching for all the data,
which means you can search quicker and speed upyour time to real time insights.
That's why a lot of companies have been movinginto data lakes because storage is getting much
cheaper.
In fact, today, we're seeing over 98% of thecost to run an application is the compute of
(28:47):
running the insights, not the fetchingtransmission or storage of the data.
That's less than 2% of the cost.
And that goes back, Joel, to your point earlierthat managing those cloud compute costs is the
most important way for startups to scaleeffectively, especially with their venture
capital dollars.
(29:08):
Yeah.
No.
That's really helpful.
We got around 10 and we gotta run soon.
So maybe you can we'll take a few questions ifpeople have it, but maybe you can tell us a
little bit about maybe you could just sharesome advice that you have for people that are
looking to pivot into VC and what worked foryou and what you would advise people to do if
they're trying to break in and if they'rehaving some trouble doing that.
(29:33):
So for me, what I saw getting into the venturecapital industry was first you wanna get your
your hands and your feet wet, which is byexploring what does VC mean.
So if you're someone who comes from finance, itit is different than other parts of the
industry like private equity and other areas.
You do want to make sure you can try a coupleinvestments and and start getting involved with
(29:58):
diligence and get involved with different angelnetworks.
When you get started also, if it's your firsttime in the industry, say, start slow.
You know, put in some small checks or or wait awhile, sit the sidelines, see what the deals
look like.
There's gonna be plenty of opportunities comingahead.
And as we're seeing in 2022, there's no sign ofventure slowing down as family offices,
(30:20):
endowments, and fund the funds are continuingto double down on their investments in the
venture capital space.
If you're someone who doesn't come from VC yet,the best way is to start getting involved with
a network.
Roll up your sleeves.
Hey.
I'll help you with technical due diligence.
If you're someone who's focused on health careor finance or on data or infrastructure, lend
(30:43):
that to a team, and you can see a lot ofbenefits by partnering with those venture
networks.
And when you get started, it's also importantto identify what are the goals you wanna do for
being in the venture industry.
Does that mean you wanna launch your ownsyndicate or fund?
Do you wanna join a fund?
Do you want to be part of a university networkof investors?
(31:04):
There's a lot of opportunities to consider.
So you wanna make sure the lines with yourmission purpose and goals.
Yeah, and I think a syndicate is a great way tostart as well.
You you can kind of go on AngelList and alsojust get some early stage signs of a track
record from sourcing deals and then gettinginsights from the community as well.
(31:25):
And I think to your point, and I've done thistoo, I built the community first, right?
So, I think building a really strong highquality community, and that means vetting is
involved.
That keeps a tight circle of just high qualitypeople.
And I feel that's for both of us, feel likethat's helped our networks compound with just
other great people to add to the circle.
(31:46):
So, I think that's a huge thing too.
One thing that I would also say is justbuilding in public.
So, if you're working on a really cool projector prototype, maybe just share that with people
and see what people think.
Would you recommend writing content and blogs?
I've seen that as a pretty common piece ofadvice to some people kind of write an
(32:07):
investment thesis if they want to eventually belike an intern for you or something or someone
else.
They could probably write some type of blogpiece that's really, really tied to what that
investor is kind of investing in and maybe evensurface some interesting deals too, I'd say,
I I absolutely encourage you to build inpublic.
(32:28):
One thing that we've done in public is launchthe data power data writer, which is now
becoming the data pledge.
So you can think, for example, of Melinda'sgiving pledge.
And the same thing here with the data pledge iscommitting to ethical and responsible use of
data on the cap table with investors and thecommunity.
And when I first launched the working document,I shared this with over 50 experts and
(32:53):
luminaries in the field to say, hey.
What's your thoughts?
Do you have draft edits?
Do you want to contribute or collaborate here?
And it evolved.
It changed, at least five times to its currentsteady state and will continue over time.
So I I do encourage you to enlist thecommunity.
And with all the projects I do, I do enlist,our LPs because, I think the best, LPs on the
(33:14):
cap table are smart LPs who provide value add,and that can mean customer introductions,
hiring, trends on the industry that they'reseeing, different ways to partner together.
And we see that that when you ask questions,often you'll get back feedback, you'll get
advice, you'll get answers, which is veryhelpful when you scale.
(33:34):
Yeah, totally agree.
Well, wanna get you out of here for your hardstop.
So if nobody else has any questions, feel freeto chime in if you have a last minute question.
The question I always ask at the end is if youhave any advice from maybe a friend or a mentor
just on just life as a whole.
So any feedback, it could be about your career,it could be about relationships, anything that
(33:57):
you have that kinda sticks out from a mentor ora friend or a family member.
I think the biggest takeaway I got from one ofmy mentors who's an adviser to Data Power
Ventures Fund.
So Chris Sanchez, he also launched the DataOath, and he partnered with me for launching a
data writer, data pledge.
And what he told me was very interesting.
(34:18):
He said, consider that what you're building isbest in class.
Consider you're building the next black stone,and that what you're building is ready and that
you're good enough to launch and scale what youwant to do, and that you don't need to be an
expert in everything.
Right?
You can align yourself with other advisers andventure partners to help you scale and see the
(34:39):
industry as a whole.
And when you shared that advice to me, that wasvery thoughtful because it helped me accelerate
my growth, be willing to not be aperfectionist, and to continue making updates
to our product offerings, to our platform.
And by doing that a lot faster, it's going backto your point, Joel, build in public.
(35:00):
And then when you're seen, you also get thisgreat feedback.
So great minds think alike.
Yeah.
Well, thank you so much.
Well, guys, if you have a question, feel freeto yell it out now within the next five
seconds, if not, and feel free to jump in ifyou do.
If not, hey David, I wanna let you go because Iknow you got a hard stop.
(35:21):
So shoot me a note later and hey, thanks againfor popping in at such a last minute and good
luck with everything.
Catch up soon.
Thanks everyone.
Thanks for having me.
Happy to answer any questions and then make ita great
Lot of these big data players who have scaledare now going public, but that doesn't mean
(00:05):
that the war has been won.
In fact, it's only the early days because mythesis is just how we looked at the twenty tens
where every company is a software or technologycompany.
Now it's the world where every company is adata company.
Welcome to The Investor, a podcast where I,Joel Palafinkel, your host, dives deep into the
(00:30):
minds of the world's most influentialinstitutional investors.
In each episode, we sit down with an investorto hear about their journeys and how global
markets are driving capital allocation.
So join us on this journey as we explore theseinsights.
PhD.
That works for me.
(00:53):
Less syllables.
Cool, I think we are live now.
So hey, David, really excited to have you onthe show.
David and I have just become friends over thelast probably six, seven months and excited to
have him in our fund accelerator as well.
But really just, it's interesting to see somany different funds now from my experience
(01:17):
break out into specific niches.
We've had psychedelics funds, we've hadhealthcare funds, climate change funds.
You're the first fund that I've met that has astrict focus on AI, ML and data science.
And I would say, this is something that youcould probably educate us on, really just the
trends and the exits.
Because I feel like some of these large datascience companies are the ones that are really
(01:40):
getting acquired by massive, probablyinstitutions and then larger conglomerates like
Amazon.
So I'm excited to kind of learn a little bitabout that as far as the trends and exits and
then just the space.
And I think another big thing that would behelpful even for me and the people here is to
just unpack data science.
So there's AI, ML, there's explainable AI,that's another hot topic.
(02:06):
But before we nerd out on all that stuff, whydon't we talk about you, David?
I know you're from Florida like myself, solet's go back to there.
Let's go back to where you grew up, where youstudied and how that helped you pivot into VC.
Yeah.
Thanks, Joel.
Thanks so much for having me on the show andproud Floridian here.
(02:26):
I like to always say, go Gators, the Universityof Florida alumni.
And for me, it does go back to the collegedays.
In fact, I was studying actuarial science,finance and information systems, wanted to go
hardcore into data.
And I discovered in my first internship where Iworked for Aflac doing loss experience
(02:48):
monitoring.
And and that means to teach me like I'm five,that that means actually predicting what the
price should be for your insurance, whetherthat's car insurance or health insurance.
And I discovered at that time that the cloudwas emerging and that no longer did working
with algorithms have to be in Excelspreadsheets, but could be in large distributed
(03:13):
systems and with programming languages likeSQL, Python, and R, which tapped me into
wanting to evolve into more of the data field.
And so I've spent now over the past decade,working for both the big Fortune five hundreds
as well as start ups in all things data.
I led and scaled data analytics and dashboardsand products, beyond Aflac to, Deutsche Bank,
(03:40):
Citigroup, and ADP, and that was all inFlorida.
I worked for some great organizations there andand then, of course, decided to take my startup
plunge, into the startup ecosystem when I movedto New York City about eight years ago.
And tell us about moving to New York because Ican probably really relate.
(04:01):
It's a different ecosystem, it's a differentenvironment as a whole.
Mean, people in general might call us too laidback or just say that we're a little more less
hyperactive as the typical New Yorkers.
So, tell me about just your thoughts on NewYork and what made you think about New York and
(04:26):
then, you know, just what you experienced whenyou first moved there from Miami.
Sure.
So I did the reverse migration.
You see many people, especially in the pandemicor or later in their life where they moved to
Florida, and I decided that moved to New York.
And for me, was fascinating because all thecompanies I worked for, Deutsche Bank,
Citigroup, ADP, all have New Jersey and NewYork offices.
(04:49):
So over my first part of my career, was goingtwice a year up to the offices and saying, what
is this ecosystem?
What am I missing out on?
In fact, I also considered moving to SiliconValley, and I spent a few months living in the
Bay Area saying, I'm thinking about SF or NewYork.
And I landed on New York for a few primereasons.
(05:09):
The first is the diversity of the ecosystem.
You have every industry imaginable in New Yorkfrom business, technology, fashion, and
finance, and that creates a wealth of newstartup ecosystem, innovation, and ideas.
And you also have a wealth of individualsworking in different careers.
You have the sales development reps and accountexecutives.
(05:31):
You have the engineers, the CEOs, and thefinance team members.
In Silicon Valley, I only saw engineers.
And for me, it was important to align close tothe business, which led me to make it to New
York.
And I think one of the best things about beingin New York is the fast moving pace of the
lifestyle.
I've traveled to different cities around theworld like Singapore, Taipei, Paris, London,
(05:55):
and nowhere is it as fast paced in New York.
So today when I work with founders andinvestors, I'd say there's nothing like having
a New Yorker on your cap table.
Yeah.
No.
I totally agree.
And then tell me about the the startup you'reworking for and what your role was there.
So today I work I I continued going into thestartup ecosystem once I moved to New York.
(06:17):
Mhmm.
And two of the startups I helped scale wereGeneral Assembly and Galvanize, which were
leaders in the EdTech ecosystem.
And I was scaling their enterprise data scienceteams, which meant I was supporting Fortune 500
clients like Charles Schwab, Bloomberg,Refinitiv, Invesco, USAA.
And after both of those companies wereacquired, I joined a little over a year and a
(06:39):
half ago, SingleStore.
And SingleStore is a database startup.
It was Y Combinator 2011.
We're backed by Google Ventures, Insight,Cosla, Dell, and HPE.
And today, eleven years into its journey, inthe last eighteen months, SingleStore raised a
series e and series f finance and is continuingto scale on its pathway to being beyond a
(07:03):
unicorn.
We're on that that territory, but primarilyhaving the passion and purpose to creating an
industry where any company can be a datacompany.
Where companies no longer have to think aboutwhere their data is warehoused or what data
lake the data is being managed on, but thateveryone can build for real time applications,
(07:24):
for data intensive applications, and for fastanalytics.
And so my role being there has been scaling ourtechnical enablement that's been with clients,
partners and employees, as well as looking forstrategic initiatives for partnerships.
You make a, you know, and you're a perfectexample.
I mean, there's a lot of people that havebecome an investor with the industry experience
(07:47):
that they've had.
So do you think that's a good starting pointfor a lot of people if they work in healthcare,
if they're working like healthcare IT to maybejoin like a digital health fund?
Or do you think it's also good for people tojust kind of jump in to other industries as
well?
And it sounds like for you, I mean, that's kindof been your niche, right?
You really built a lot of great enterprise andindustry experience in data.
So it just kind of makes it a natural fit tojust, you've already understood all the
(08:09):
problems in the business models to to just jumpin and directly be on the investor side.
Yeah.
I really appreciate what you shared earlier,Joel, which is about niching down into a
vertical.
And for me, it's just well, my whole career hasbeen around data ML AI.
So it made sense.
Take what I know and apply that to a diligenceand to mentorship and advisory for startups,
(08:34):
which led into investments.
So I do think, if you're someone who's emergingas a fund manager, if you're getting into the
angel investing industry as well, start withwhat you know.
It is a vast industry.
When we started, our thesis was about beingdata driven, but we realized very quickly that
was too general, and so then we narrow downinto data intensive apps, real time insight,
(08:59):
data developer tools by focusing on data power,those infrastructure layers, those insight
layers, and those prediction layers of theeconomy.
I think that's really helpful when you're a newinvestor because otherwise, you're seeing so
many opportunities, and it's tough for you tobuild your mental models.
You wanna build the best mental models so youcan decide how do you make investments.
(09:22):
And I think that's based on either theknowledge you know or the team you surround
yourself with.
I've also heard, you know, I just wanna make acomment to your earlier point.
So with the niche funds, I've I've also seenthat be really attractive to LPs because LPs do
have different interests.
They're interested in Web3, they're interestedin quantum computing, they're interested in
(09:43):
psychedelics, but they don't have the time tosource these deals and they don't have the
network and they may know, they may not havethe technical chops.
A lot of them do because they're pastentrepreneurs, but a lot of them just may not
have the technical chops to due diligence.
And one thing that you and I talked about,which helped you, I think, form a really
interesting partnership with someone was yourability to kind of do technical diligence with
(10:05):
someone else.
So I think that's an interesting point.
LPs want exposure to an exciting sector that isreally hot, but they just don't have the
bandwidth to go out and source all those dealsindividually.
And my favorite analogy is, you can either buyTesla ad hoc or you could just buy it index
with like a bunch of the FANG companies in acertain theme or something like that.
(10:29):
So I just wanted to make that comment, but youstarted doing this.
I want you to go a little deeper.
Can you just unpack the whole universe for usfor just data science and how the investment
activity is broken out?
So we talked a couple of things we talkedabout, right?
Explainable AI, AIML.
Can you just kind of give us a high leveloverview of just the whole landscape and then
(10:50):
and then where you play?
You've talked about like the data power.
So maybe just kind of elaborate a little morefor our education.
So when we think about software and data, thebig trend to look at is the emergence of the
software engineering industry.
And so software engineering had its rise tofame in the early two thousand.
(11:11):
I would say that even went to, like, 2010 and2015 with software automation and all the tools
so that every engineer has a toolkit where theycan build the best world class systems.
What we're seeing with data is that has notexisted until now.
The twenty twenties are known as the decade ofdata.
This is where the modern data stack's beingbuilt.
(11:33):
New developer tools, new infrastructure, newsystems so that data scientists, ML engineers,
and AI specialists alike can have their owntoolkit to also build and scale enterprise
grade data systems.
And to do that, it doesn't only require thetools, it requires the systems and having them
talk to each other.
(11:53):
So we've also been seeing as a result of that,some of the early players like the single
stores who've been around since 2011 to 2015focused on data are now going through m and a,
getting acquired, and or IPO ing.
We've seen companies like Datadog go on to thepublic markets, MongoDB, Snowflake.
(12:14):
A lot of these big data players who have scaledare now going public, but that doesn't mean
that the war has been won.
In fact, it's only the early days because mythesis is just how we looked at the 20 tens
where every company is a software or technologycompany.
Now it's the world where every company is adata company.
(12:38):
That's exciting.
And what are some of the the segments thatyou're most excited about kinda unpacking that
with the with the data, focus, I guess?
So I look at the data industry as four layers.
The first layer is data collection.
And data collection has typically been a verymanual process with labeling and aggregating
(13:01):
the datasets and then storing them.
So that's beginning to automate.
That's a very exciting space.
It's growing a lot.
We've seen Labelbox, Scale AI, and otherplayers become unicorns in that space.
We even see newer companies like Roboflow,which are helping automate some of the labeling
for computer vision and NLP systems.
(13:21):
Then we have layer two, and layer two is reallyabout the real time insights.
It's having tools that are dashboards orspecific for different industries, whether it's
health care, biomedical, or looking at securityin in the cyber industry and having these tools
so a company can say, I can buy this tool andhave data layer integrate with the APIs and get
(13:46):
real time insights without having an army ofdata scientists and data analysts to generate
these insights.
So that's layer two, which we're seeing a lotof platforms emerge and scale in that space.
Layer three is where we start seeing thepredictive insights.
That's the traditional machine learning that'sbeen told of, if I apply for a loan today, do I
(14:07):
get approved or do I get rejected?
And looking at those insights, in real time isbeneficial for industries to not have to have a
delay of taking a few days for results, butinstead having instantaneous data intensive
apps.
And so we're seeing that today in underwriting.
We're seeing that in credit reporting.
We're seeing that in these new, ML powered appsthat support, you know, your need to be more
(14:34):
successful at enterprises.
And then the fourth layer is about AI.
So most companies today talk about we're AIfirst, but if you don't get the other three
layers right, then the fourth layer won't getthere just yet.
And so the fourth layer of being AI poweredmeans you're using this modern AI stack, which
(14:54):
could be like OpenAI's GPT three to generatenatural language processing text or create
generative images or create systems that selflearn and heal on themselves.
The challenge when companies talk about beingAI powered is it's the whole ecosystem of data,
ML, and AI, not AI alone.
(15:15):
And so I'm really excited about all four ofthose layers, and we looked at all of them with
our investment thesis and mathematics.
It's really exciting, and that that was reallygood learning for me.
Where do you think quantum will play in all ofthis?
I guess will quantum have a because from mylearnings, I'd love for you to clarify or help
me with this, some of the bigger applicationswith quantum are like material sciences,
(15:41):
infrastructure.
There's also some chemical applications aswell.
So, I guess there's applications for all ofthese that you mentioned, right?
The collection, the automation, the predictiveand AI power stacks, but how will quantum
evolve into kind of this whole ecosystem?
(16:02):
I'm very excited for quantum and I think it isstill the early days of quantum.
I do have several good friends who work in thatspace.
So we routinely chat about the trends andchanges that we're seeing in the quantum
industry.
Just this month, I've looked at this reallyexciting quantum startup that's using quantum
entanglement to replace global GPS withsatellites to get better real time data for
(16:28):
clocks and movement of trains around the world.
And when you you hear this story for the firsttime, most people say, is this really a
problem?
But the truth is there is.
There was cases where thirteen microsecondswere off between some of these, travels, which
caused flight airline, disruptions andinterruptions.
You even have trading systems where there'sthirteen microseconds that can mean the
(16:50):
difference between a profitable or a lossdriven month.
So I think there's a lot of opportunity inquantum, and we've seen that with the cubits.
We've seen our systems scale to beyond a 100cubits.
I I think we need to get to that thousand milemarker to start seeing some benefits just like
we saw in the AI space with the scale of GPUsand now with ASICs in in the blockchain space.
(17:15):
So Quantum will have its day, and and therehave been some unicorn IPOs there as well.
I just think it's still the early days, and weneed to find commercially viable business cases
to work with Quantum.
Yeah, no, it's really exciting.
What are some pieces of advice that you wouldhave for people if they're interested in being
data science investors for sourcing andscreening?
(17:37):
Are there a lot of these, are they still in theacademic ecosystem?
Because I know a lot of the quantum deals andthe technology is still in like tech
transformation.
Some of the events that I went to in the pasthad a lot of university presence kind of doing
a lot of research, but is that the similarspace to kind of meet a lot of these data
startups?
(17:58):
Are there great accelerators that areincubating these startups and technologies?
I guess what are the best avenues to kind oflearn about these opportunities?
Yeah.
So I do partner with a lot of accelerators andsome of my favorite ones, of course, are
Techstars being in New York City.
I've seen a lot of companies emerge.
So they're building developer tools around dataand ML in the Techstars ecosystem.
(18:23):
Of course, there's a lot of ways that you candiscover the market because there are so many
data startups today.
In fact, my mission is over the next, ten yearsto accelerate a thousand data powered startups.
So I think we're gonna continue to see morefounders and more, executives and engineers
alike move into the data ecosystem.
(18:44):
But when all things are considered, if you'relooking to invest, I I think it goes back to
your point earlier, Joel, that there's a coupleways.
Right?
You can go deal by deal or you could do thatindex portfolio.
And that's what we see with a fund like DataPower Ventures Mhmm.
That we've had LPs come in really for a couplereasons.
First, they're either the engineers or businessexecutives who build or scale technology, and
(19:05):
they understand that data is the next wave thatwe're seeing in industrial revolution.
Or secondly, they want that exposure.
They want that education.
They're a finance executive who doesn't knowmuch about data ML AI, but they also believe
that, hey.
This is something that I don't wanna miss outon.
I want the education information rights and tohave also some upside in my portfolio.
(19:27):
Yeah.
No.
I totally agree.
I think those are all, good reasons to becomeLPs.
And it's also just, it's great too if you canbuild a relationship with LPs and have these
focus groups and kind of keep them involvedbecause a lot of them have that industry
experience as well.
Some of them are really well known datascientists.
(19:49):
So, they may actually help you de risk some ofthe investments as well.
So, I think if you keep them engaged and kindof seek feedback, that also helps to build that
long term relationship as well.
That's just kind of what I've seen.
And I can probably assume you agree too.
Community is everything.
Absolutely.
(20:09):
Yeah.
David's got a really great, I wanted to plugthis too.
You've got a really amazing community oftechnologists as well.
You have a tech dinner that you host every oncein a while.
Is there a website where we can go to sign upfor that?
Or is that just more off the record kind ofconnecting with you?
Yeah.
So there's a few community driven events that Ido, which I can share.
(20:32):
You know, number one, with, the podcast, I dohave an AI data science podcast been running
for over three years called Humane with AI inthe middle.
So that's h u m a I n.
And we have Humain podcast where you can listento great episodes with founders from Preseed to
IPO Ventures.
Data Power is at datapower.vc.
(20:52):
So you can, of course, learn more about ourportfolio and and how we're investing and and
everything around there.
And the Tech Dinners, this is an invite onlycurated community around investors, founders,
developers in New York City.
But to get involved with the Tech Dinnerseries, you can find me on LinkedIn, and you'll
(21:13):
see my most recent saved item on my profile hasan Airtable link.
So I'm making it really hard for you to get to.
But if you're interested and excited, you cancheck it out.
And we we run these events in New York City.
We've actually brought multiple portfoliocompanies in person with LPs.
They've directly invested after meeting thefounders.
(21:35):
We've brought in near introductions who've beenhired by the portfolio companies.
So they're very strategic community drivenevents.
Yeah, I mean, I've been seeing a lot of peopleuse NFTs now to kind of have utility in the
community and also get access to events aswell.
So I wouldn't be surprised if you had a datapower NFT launching at some point in the
(21:55):
future.
Let me think, what are some other hot trends?
So I think, one thing that I've seen too, andthis probably falls into the tools, I've seen
some platforms that help to just provide costsavings.
So like, I think Cloud Admin is one of themwhere they help to kind of help you manage your
costs.
So do you think that's probably a hugeapplication and very, very high value to data
(22:19):
scientists, but do you think that's a hotindustry or there are a lot of people doing
that now?
So we've seen a few a few startups in thisDevOps automation space in the last couple of
years.
And and I agree that this is very important.
Even at single store, we face these costs everysingle day because we're a cloud first company
and our bill is, at minimum, into the hundredsof thousands of dollars a month on cloud costs.
(22:44):
And just by using some automations that we setup recently, we're now saving, at least a
$100,000 a month.
This is no joke.
Right?
And this is just by an engineer writingscripts.
So there are startups that also are, focused onthis business where they can plug in to your
cloud, AWS, GCP, Azure, and then helpunderstand think about these systems turning
(23:06):
off, turning on, or using more reserved ordedicated instances, and you can see those
changes.
I think that's very important, because one ofthe underlying factors that we've seen as a
result of the pandemic is that every company isdigital first.
You no longer have a choice.
And even in the hybrid world, that means cloudis here to stay, and a cloud first strategy is
(23:29):
required.
Yeah.
No.
I totally agree.
Where do you see data?
So we've seen a lot of these companies do likeUI automation.
So what's kind of like super deep tech for you?
Like, do you feel I'm trying to think of likesuper sci fi stuff with data.
So I think when I can think about it, I canthink about like just real time movements of
(23:50):
your body and like maybe predicting.
What I think would be really interesting is ifyou could, and they're probably already doing
this, but based on your family history andbased on kind of your eating habits in real
time, like imagine if they could just detectyour lifespan.
If you still eat pizza and fried chicken likefour to five times a week, they could probably
(24:14):
already predict some outcomes because there'sso much data that's been collected.
There's probably like some parts of it too tiedto like your ethnicity and your cultural
background too.
But I feel like that would be a really hotspace for just like real time data And then
predicting kind of what your future be ifyou're continuing to do what you're doing.
(24:34):
There could be results.
There was a company that I that I met three orfour years ago.
And they were trying to do that.
I don't know if they got the traction, but theyclaim to be able to detect within high
accuracy, like when you would get a heartattack based on your activity.
But I don't if you've been seeing anything likethat in the health care space, but that's got
to be a huge insurance play as well with youcoming from like Aflac as well, maybe.
(25:00):
There there's so much opportunity in the deeptech sci fi space that is really fascinating.
And, companies we've looked at, that we haven'tparticipated in, but we've we've explored, one
includes a a satellite company that's producingsatellites at one tenth the cost.
But what's great is the satellites are trackingdata about where objects are in space to then
(25:21):
provide the insights for, space trash removaland clean up.
So that's a really, deep tech play.
Oh, wow.
Another one is a a sensor company that installsin warehouses these special sensors to see
where any movement occurs anytime based onvibrations and waves occurring in those
(25:41):
warehouses.
And this is a great company because, you know,Boeing and Maersk and others have already
committed MOUs to them.
So we're really looking at that kind of deeptech to see all these motions.
So I think those are a couple really uniqueones.
One that's more down to earth on, literallyspeaking on the health side, that we're
finishing our investment in, is is focusedaround accelerating research and development
(26:06):
for biopharmaceutical companies, and thatcompany is called Applied XL.
And so Applied XL, is led by, Francesco, who'sthe former head of r and d at the Wall Street
Journal.
Really exciting startup that we're finishing toround out with Data Power Ventures.
And and that company actually, is is alreadybeing led by some really exciting, investors
(26:28):
all in New York City.
So we do love to support New York founders.
We've actually have five of our portfoliocompanies are based out of New York.
So there's, I think a lot of movement we'reseeing today in the entire data ecosystem.
Yeah.
And then I think one thing that I've seen inthe last probably five or six years is just
kinda advances in how you're housing the dataand storing them and categorizing them.
(26:49):
So maybe you can share a little bit of insighton like just some of the enterprise workflows.
Like now they have the concept of a data lakewhere you're kind of taking the data
categorizing in this lake where you can kind ofuse it whenever you need.
But can you educate us a little more on kind ofsome terminology that like if you work at a big
data company, some things to think about, ifyou're like a data scientist on the enterprise
(27:12):
side, people may not know about it because theydon't work at these companies.
So but I think you might have a little moreinsight into just kind of the industry because
there's just a lot of internal workflows thatpeople need to think about.
At sales store so we we call it actually threetiered storage.
So this is our our name of the technology,which is similar to data lake.
(27:34):
It means that, one, you actually have the datastructures that you house the data in.
For us, that's row stores and column stores.
For other companies, those are data frames.
Those are CSV spreadsheets.
Those are SQL tables.
Then below that, we actually have the storageon your hardware.
So that could be, in this case for us, theseare the compute units, which are Linux AMI
(27:57):
machines, and they have their own, databasethat's sharded or distributed, into storage on
these machines.
And then where the data lake appears, we see itwith Databricks, where they have their
universal data lake and their Delta Lakesystem.
And then for single store, it's where we havewhat's called bottomless or unlimited storage.
So pretty much, it's like you have this oceanof data, and you can dip into it with a fishing
(28:22):
net and grab the data that you need when youneed it without searching for all the data,
which means you can search quicker and speed upyour time to real time insights.
That's why a lot of companies have been movinginto data lakes because storage is getting much
cheaper.
In fact, today, we're seeing over 98% of thecost to run an application is the compute of
(28:47):
running the insights, not the fetchingtransmission or storage of the data.
That's less than 2% of the cost.
And that goes back, Joel, to your point earlierthat managing those cloud compute costs is the
most important way for startups to scaleeffectively, especially with their venture
capital dollars.
(29:08):
Yeah.
No.
That's really helpful.
We got around 10 and we gotta run soon.
So maybe you can we'll take a few questions ifpeople have it, but maybe you can tell us a
little bit about maybe you could just sharesome advice that you have for people that are
looking to pivot into VC and what worked foryou and what you would advise people to do if
they're trying to break in and if they'rehaving some trouble doing that.
(29:33):
So for me, what I saw getting into the venturecapital industry was first you wanna get your
your hands and your feet wet, which is byexploring what does VC mean.
So if you're someone who comes from finance, itit is different than other parts of the
industry like private equity and other areas.
You do want to make sure you can try a coupleinvestments and and start getting involved with
(29:58):
diligence and get involved with different angelnetworks.
When you get started also, if it's your firsttime in the industry, say, start slow.
You know, put in some small checks or or wait awhile, sit the sidelines, see what the deals
look like.
There's gonna be plenty of opportunities comingahead.
And as we're seeing in 2022, there's no sign ofventure slowing down as family offices,
(30:20):
endowments, and fund the funds are continuingto double down on their investments in the
venture capital space.
If you're someone who doesn't come from VC yet,the best way is to start getting involved with
a network.
Roll up your sleeves.
Hey.
I'll help you with technical due diligence.
If you're someone who's focused on health careor finance or on data or infrastructure, lend
(30:43):
that to a team, and you can see a lot ofbenefits by partnering with those venture
networks.
And when you get started, it's also importantto identify what are the goals you wanna do for
being in the venture industry.
Does that mean you wanna launch your ownsyndicate or fund?
Do you wanna join a fund?
Do you want to be part of a university networkof investors?
(31:04):
There's a lot of opportunities to consider.
So you wanna make sure the lines with yourmission purpose and goals.
Yeah, and I think a syndicate is a great way tostart as well.
You you can kind of go on AngelList and alsojust get some early stage signs of a track
record from sourcing deals and then gettinginsights from the community as well.
(31:25):
And I think to your point, and I've done thistoo, I built the community first, right?
So, I think building a really strong highquality community, and that means vetting is
involved.
That keeps a tight circle of just high qualitypeople.
And I feel that's for both of us, feel likethat's helped our networks compound with just
other great people to add to the circle.
(31:46):
So, I think that's a huge thing too.
One thing that I would also say is justbuilding in public.
So, if you're working on a really cool projector prototype, maybe just share that with people
and see what people think.
Would you recommend writing content and blogs?
I've seen that as a pretty common piece ofadvice to some people kind of write an
(32:07):
investment thesis if they want to eventually belike an intern for you or something or someone
else.
They could probably write some type of blogpiece that's really, really tied to what that
investor is kind of investing in and maybe evensurface some interesting deals too, I'd say,
I I absolutely encourage you to build inpublic.
(32:28):
One thing that we've done in public is launchthe data power data writer, which is now
becoming the data pledge.
So you can think, for example, of Melinda'sgiving pledge.
And the same thing here with the data pledge iscommitting to ethical and responsible use of
data on the cap table with investors and thecommunity.
And when I first launched the working document,I shared this with over 50 experts and
(32:53):
luminaries in the field to say, hey.
What's your thoughts?
Do you have draft edits?
Do you want to contribute or collaborate here?
And it evolved.
It changed, at least five times to its currentsteady state and will continue over time.
So I I do encourage you to enlist thecommunity.
And with all the projects I do, I do enlist,our LPs because, I think the best, LPs on the
(33:14):
cap table are smart LPs who provide value add,and that can mean customer introductions,
hiring, trends on the industry that they'reseeing, different ways to partner together.
And we see that that when you ask questions,often you'll get back feedback, you'll get
advice, you'll get answers, which is veryhelpful when you scale.
(33:34):
Yeah, totally agree.
Well, wanna get you out of here for your hardstop.
So if nobody else has any questions, feel freeto chime in if you have a last minute question.
The question I always ask at the end is if youhave any advice from maybe a friend or a mentor
just on just life as a whole.
So any feedback, it could be about your career,it could be about relationships, anything that
(33:57):
you have that kinda sticks out from a mentor ora friend or a family member.
I think the biggest takeaway I got from one ofmy mentors who's an adviser to Data Power
Ventures Fund.
So Chris Sanchez, he also launched the DataOath, and he partnered with me for launching a
data writer, data pledge.
And what he told me was very interesting.
(34:18):
He said, consider that what you're building isbest in class.
Consider you're building the next black stone,and that what you're building is ready and that
you're good enough to launch and scale what youwant to do, and that you don't need to be an
expert in everything.
Right?
You can align yourself with other advisers andventure partners to help you scale and see the
(34:39):
industry as a whole.
And when you shared that advice to me, that wasvery thoughtful because it helped me accelerate
my growth, be willing to not be aperfectionist, and to continue making updates
to our product offerings, to our platform.
And by doing that a lot faster, it's going backto your point, Joel, build in public.
(35:00):
And then when you're seen, you also get thisgreat feedback.
So great minds think alike.
Yeah.
Well, thank you so much.
Well, guys, if you have a question, feel freeto yell it out now within the next five
seconds, if not, and feel free to jump in ifyou do.
If not, hey David, I wanna let you go because Iknow you got a hard stop.
(35:21):
So shoot me a note later and hey, thanks againfor popping in at such a last minute and good
luck with everything.
Catch up soon.
Thanks everyone.
Thanks for having me.
Happy to answer any questions and then make ita great
Popular Podcasts
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.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Special Summer Offer: Exclusively on Apple Podcasts, try our Dateline Premium subscription completely free for one month! With Dateline Premium, you get every episode ad-free plus exclusive bonus content.