June 18, 2025 • 41 mins
It seems like we are bombarded by news about millions of dollars pouring into AI startups, which have crazy valuations. In this episode, Chris and Dan dive deep into the highs, lows, and hard choices behind funding an AI startup. They explore early bootstrapping, the transition to venture capital, and what it’s like to trade in code commits for investor decks.
Featuring:
Links:
Sponsors:
- Miro: Feeling overwhelmed by AI? Miro brings clarity by combining human creativity with intelligent tools to help teams get great work done. Learn more at miro.com.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Jerod (00:04):
Welcome to the Practical AI podcast, where we break down
the real world applications ofartificial intelligence and how
it's shaping the way we live,work, and create. Our goal is to
help make AI technologypractical, productive, and
accessible to everyone. Whetheryou're a developer, business
leader, or just curious aboutthe tech behind the buzz, you're
(00:24):
in the right place. Be sure toconnect with us on LinkedIn, X,
or Blue Sky to stay up to datewith episode drops, behind the
scenes content, and AI insights.You can learn more at
practicalai.fm.
Now, onto the show.
Daniel (00:48):
Welcome to another fully connected episode of the
Practical AI Podcast. In thesefully connected episodes, it's
just Chris and I, and we spendsome time catching up on AI news
and and maybe some other topicsthat are just on our mind and
hopefully share some things withyou that will be useful and
(01:09):
educational as you explore theAI world and and your own, your
own work with AI. I'm DanielWitenack. I am CEO at Prediction
Guard, and I am joined as alwaysby my cohost, Chris Benson, who
is a principal AI researchengineer at Lockheed Martin. How
are doing, Chris?
Chris (01:28):
Doing well today, Daniel. So much is happening. Looking
forward to our conversationtoday. Yeah. So much is
happening.
Daniel (01:35):
It's summer is upon us, which is is great. If my
allergies are any indication,then then things are things are
changing. But, yeah, there's athere's a lot of interesting
things changing and shiftshappening in in the AI world.
One of the things that I'm I'mconstantly just I see all the
(01:55):
time in the news, we we have a ateam, you know, a channel in our
in our Slack, just a randomchannel where people share news
stories. And there's alwayssomething about, you know, AI
startups, either good or bad ofof AI startups.
Actually, this this last week,one of the one of the things
(02:16):
that was shared was this, youknow, hyped AI startup that that
raised a bunch of money, wasvalued at 1,500,000,000.0,
Builder dot ai or builder.ai.And I think, basically, it it
collapsed. There were sort ofthese these I I don't know
everything about it, but I sawsort of posts about the AI not
(02:41):
being AI, but actually being,you know, Indian coders or
developers. So it was like a abuilder platform, like a low
code, you know, I'm gonna buildan application, but actually,
there's some indication thatthere was actually just people
building building behind thebackground. I I don't know the
full story behind that or ifthat's even true.
(03:01):
But regardless of the sort ofcollapse, I don't know if you're
seeing a similar flurry of allsorts of discussions of AI
companies, good or bad.
Chris (03:11):
Yeah. And and, you know, there's there's kind of never
ending talk of AI bubbles and,you know, fundings up and down.
And and I know you as a as astartup founder yourself have
have a lot of interesting,experiences with the funding
world and stuff like that. Maybemaybe get you to share some of
(03:31):
that today.
Daniel (03:32):
Yeah. Yeah. Sure. A little bit of the what what is
it what is it actually like inthe trenches, to raise funding
for for an AI startup? What whatperceptions are there?
What questions come up? Youknow, what stages are there? I'm
I'm just looking at I search forAI funding startup in in Google
(03:54):
News, and it's just sort of a aconstant feed. The ones that are
popping up as we are recordingthis episode, legal startup
Harvey AI raising funding at a$5,000,000,000 valuation,
Decagon to raise a 100,000,000at 1,500,000,000.0 valuation,
Samaya AI, gets 43,000,000.Perplexity is raising around
(04:21):
bringing in 500,000,000 at a14,000,000,000 valuation.
Yeah. It may be even useful forsome people out there that
aren't as much in the startupworld to even just parse through
some of what this actually eveneven means, like some of these
(04:42):
terms. So maybe that would be agood set of first questions to
answer is what does it actuallymean to raise funding regardless
of whether you're an AI startupor not?
Chris (04:53):
And maybe even before we, you know, just as a part of
that, even what some of thatbasic, terminology means, you
know, where you're where you'reraising a certain amount at a
certain valuation. Yeah. Forthose who have not been exposed
to to the finance side of thingsin the past, you know, and and
what is the how does the thelevel of valuation affect your
funding and affects control,things like that? So
Daniel (05:16):
Yeah. Yeah. Well, there's I I have to admit
myself, you know, learning a lotabout this in the in the past
year as as we've gone throughthis process. Lots of
interesting, strange, you know,terms and and processes and and
all of that. It's a whole worldin and of itself.
But yeah, so just if you thinkabout building a company, you
(05:40):
know, you're gonna build an AIcompany. Typically that company
might be fall into a fewdifferent categories, right? So
maybe you're an AI company thatyou want to provide services to
others. Like you wanna help thembuild AI things. This would be
like a service provider type ofcompany, a consultancy, that
(06:05):
sort of thing.
Typically, when you're hearingabout these companies raising
venture capital, which we'lltalk about that here in a
second, they're not reallyfocused on these sort of service
provider companies. There isactually a huge market for that
right now. Actually, I think, Iwas on a call with someone and
it's like, you know, OpenAIisn't the one really making the
(06:27):
the most money off of of AI.It's like Accenture. Like, the
the consultancies of the worldare are just making a a huge
amount of money off of AIbecause people, you know, don't
know how to how to build thesethings.
They they don't know how toadopt things, etcetera. So it is
a really good space to be in.But in in another category, your
(06:48):
AI startup might be a a productbased company. So you have a,
you know, an AI agent platformthat you're building for
petroleum engineering, or youhave a healthcare assistant
company that you're building, oryou have a developer tools
company that has developer toolsfor working with AI models or
(07:09):
platform or infrastructuresoftware. In our case for
Prediction Guard, we're more onthat infrastructure system side,
right?
But all of these are kind ofproducts. Now, if you then kind
of subdivide those products,there's a set of products that
are probably somewhat niche.Like you have an AI product, but
(07:35):
the number of people that aregonna use it, if you wanna think
about it, the market for that AIproduct is fairly small, right?
So maybe I have, I don't know,I'm thinking off the top of my
head, but I have a an AI tool oran AI product for, pug owners. I
(07:57):
had a I had a pug dog for awhile.
So I I have an AI product forpug owners, and it helps, you
know, answer questions aboutshedding or about great things
to do with your pug. Now maybe II don't know how many pug owners
there are in the world, butmaybe that's not, like, the the
biggest market that you couldthink of in terms of the amount
(08:18):
of money you could make off ofsuch a thing. I don't know if
you agree or not, Chris.
Chris (08:22):
Oh, I don't know. There's lots of pug owners out there.
Daniel (08:25):
Very true. Very true. There may be those products and
you could kind of think aboutthose if you want, like some
people might call these sort oflifestyle type of businesses
where you could actually, youcould probably make a reasonable
amount of money. Now, I don'tknow if the Pug AI would make a
(08:47):
reasonable amount of money, butyou could make a good living for
yourself, maybe.
Chris (08:52):
It it would be. So I think kind of to the point,
there's an addressable marketthat that you would that your
startup would need to be wouldneed to be working towards.
Daniel (09:03):
Correct. And
Chris (09:04):
it needs to be an addressable market that makes
what you're doing worthwhile.
Daniel (09:07):
Correct. Yes. And and sometimes there's sort of two
there's really two things youcould determine once you look at
the market that is available forwhat you're building. On one
side, there could be a marketand that market could be big
enough to support your business,maybe even employees pay you a
paycheck. Right?
(09:28):
But it's not, it's never goingto be a large market in the
sense of billions of dollars. Soreally, you know, billions or
tens of billions or hundreds ofbillions of dollars. Right? Then
there's another set of productsthat would have a larger market.
So, you know, just by way ofexample, you look at certain
(09:52):
startups from the past thathave, you know, captured a very
large market, whether that be,you know, Uber or Airbnb, I
guess those are consumer relatedthings.
But then, you know, there's allsorts of things in the computing
and B2B side. So you look atsomething like Docker or, you
(10:14):
know, whatever these kind oflarger, there's a huge market
for some of these types ofproducts. So if your AI product
doesn't just have a market of,you know, some millions of
dollars that could support yourlifestyle, but it's, you know,
tens of billions of dollars orhundreds of billions of dollars
(10:34):
of market out there, then thatkind of categorizes you in the
potential spot where you couldraise venture capital for your
startup. And so now we kind ofneed to define what that means,
which is if you're gonna raiseventure capital, basically
(10:55):
venture capitalists arecompanies that they actually
have their own investors. Sothey get together a bunch of
investors often called LPs orlimited partners that put in
money into a fund.
Let's say it's a $50,000,000fund and the VC firm then
(11:18):
allocates from that fund toinvest in a number of companies,
in quite a few companiesactually, with the hopes that
certain of these companiesactually do scale up to reach a
reasonable amount of that verylarge market, thus becoming
(11:40):
large valuable companies thatprovide a return to that
investor investor pool.
Chris (11:49):
I'm curious. One of the often before the VC stage,
you'll have angel investors,which are, you know, wealthy
individuals that are getting invery, very early Yep. Even
before the VC do and do that. Doyou have did did you guys go
that way? Or or or at least areare you seeing that in the AI
space?
Daniel (12:07):
We we didn't. In in our case, actually, we spent really
a year and a half, the firstyear to year and a half of our
business bootstrapping thebusiness, which what that that
means is basically, you know,businesses ideally make money.
And if you get an, know, if yousell enough of what you're
making and you bring money in,you can pay bills and keep
(12:31):
running your business. Right?And so that's kind of, you're
living essentially growing yourbusiness as fast as it can based
on the money that's coming inrevenue wise.
And so we did that initially.Others kind of might not be in
the place where they do that.They might not be able to get
contracts or sales early enough.And so maybe there's kind of
(12:52):
some initial money in from someangels.
Chris (12:55):
I've got a question for you on that. And that is,
there's a point there, you know,where you're bootstrapping early
on. And I remember when you weregoing through that process, and
there was a point, you know,along the way, and I don't know
where that happened, but whereyou decide you are in, you need
to go the VC route to achieveyour goals and stuff. How what's
your thinking as an AI founderabout that? You know?
(13:19):
Because Yeah. There are tradeoffs on on if you bootstrap, you
know, all the way through andgrow based on your own revenues,
you know, that gives you a acertain advantage in some ways
that, you know, you're nothaving to go through the process
of VC. As a founder, how did youhow did you really kind of
decide, you know, what the tradeoffs were and, you know, and and
(13:40):
go the way that you ended upgoing?
Daniel (13:42):
Yeah. It's a good question, and I don't think
there's any right answer to thisquestion for different people. I
think it's a question peoplehave to answer for themselves
and the business that they'rebuilding. For us, I think what
we saw was that we did have aproduct that we were creating,
(14:03):
which at least had some level offit to a real need in the
industry. And there was thechance to scale that product up
and gain market share, but onlyif we did that sort of quite
rapidly.
Because a lot of theconsolidation, I think the kind
(14:25):
of acquisition and consolidationthat will happen in the AI
market, it's not gonna happen,you know, ten or fifteen years
from now. It's gonna happen, youknow, three to five years or may
or maybe sooner from now. Right?So if you want to kind of
participate in that sort ofconsolidation, potentially be
acquired, then you really needto to kind of, scale that quite
(14:48):
quickly, which requires capital.It's not that you couldn't grow
organically, right, and stillgrow organically and have an
again, a nice lifestylebusiness, but you probably
wouldn't participate in thatkind of consolidation phase.
For us, that's ultimately whatwe decided. Is risky, but that's
(15:08):
ultimately what we decided todo.
Sponsor (15:24):
You know, with everything going on in the AI
world right now, it is easy tofeel overwhelmed. There's a lot
of noise out there. People arewondering if AI is coming for
their jobs. But Miro's latestsurvey paints a more optimistic
picture. Here's what it says.
76% of knowledge workersactually believe AI can help
their roles, not take them. Butthe problem is, over half those
(15:48):
folks still don't know when orhow to use AI. That's where
Miro's Innovation Workspacecomes into play and has become a
go to for our team. We've triedit out. We love it.
Whether we're planning anepisode, building out a new
podcast format for our client orjust brainstorming ideas, What
used to take hours across docs,whiteboards, chat threads. Now
(16:09):
it all happens in one space. Andhonestly, it's a lot of fun.
Miro makes collaborating withyour team and with your A. I.
Way less intimidating and waymore productive. The coolest
part is you don't have to be anexpert with A. I. The work
you're already doing with Miro'sCanvas becomes the prompt. You
can turn a sticky note intodiagrams, screenshots into data
(16:31):
tables and ideas into realoutcomes fast.
So help your team get great donewith Miro. Check out miro.com to
find out how that is mir0.com,miro.com and tell them you heard
about it on Practical AI.
Daniel (16:54):
Well, Chris, we talked we talked a little bit about
what we were trying to do inthis fundraising phase. We
defined a few terms like venturecapitalists. I think it might be
interesting for people to knowmaybe that haven't been exposed
to this, some of the dynamicsthat venture capitalists might
might be thinking of. Now I feelreally privileged that, that
(17:17):
that we've got some folks thathave partnered with with
Prediction Guard that arereally, with us and supporting
us in in a really, profound way,also, you know, aligning to a
lot of how we wanna build thebusiness. But if you just kind
of strip away anything but,like, the capital and return
pieces of how at least many VCsmight think, Again, going back
(17:42):
to the example that we hadbefore, there might be, you
might have a fund of 50,000,000or a $100,000,000 and you invest
in however many, let's say 30different companies.
What you're hoping for is, andyou make an investment in that
company. Essentially what you'redoing is you're, you're, the
founders are selling a piece oftheir company and you're buying
(18:04):
a piece of that company viaequity. So you're giving away
shares or equity in yourcompany, a percentage of your
company. There's various typesof mechanisms of how that
happened, whether it's just, youknow, you value the company and
you buy certain shares or youbuy the right to, you know, have
a certain percentage of thecompany later on that converts.
(18:25):
There's all sorts of differentmechanisms around how that
happens, safes and convertiblenotes and venture debt and
priced, valuation rounds, thatsort of thing.
But essentially that's whatyou're doing. And what the VC
would hope for, let's say thatwe do that with 30, you know,
Chris, you and I have$50,000,000, which I don't think
(18:48):
we do. But let's just say wedid. We don't. And we we put you
know, we divide that out across30 different companies.
We buy pieces of thesecompanies. Essentially, this is
a very risky thing to do, butwhat you're hoping is that it's
sort of like a power laws sortof thing. So you're kind of
(19:08):
hoping that one of thosecompanies just is awesome and
grows like wildfire, is acquiredfor a huge amount of money, and
that your one investment in thatone company returns you your
whole fund. $50,000,000, let'ssay. Then another three of those
(19:28):
returns your fund again.
So not as good of exits, right,but, you know, not bad. So you
basically have doubled yourmoney. And then the whole
remaining set of the companiesreturns it maybe a third time.
Right? So essentially you'vetripled your money, but most of
that has come from a very smallamount of the companies that
(19:50):
you've invested in, and you geta good, you know, you get a good
return for your investors whoput the money into to your fund.
Chris (19:58):
It's a form of diversification of risk.
Daniel (20:00):
Yeah. Essentially.
Chris (20:00):
So you're you're selecting a a bunch of different
companies and hoping that one ora a small number of them will
will far outpace Yeah. The thelarger collective.
Daniel (20:11):
Yeah. Now that's not the rule across the board, but it's
kind of generally how you couldthink of this. There's different
firms that have differentstrategies that make a smaller
number of investments or maybemake a very, very, very large
number of very smallinvestments. Right? And so
there's there's differentstrategies within that, and
that's kind of, part of that VCworld and and determining what
(20:35):
the, you know, the way that youwanna run your fund and the
thesis that you have and all ofthose sorts of things.
But that's kinda what'shappening. So if we just go back
to kind of interpreting some ofthese things. So I mentioned
that, perplexity raise, right?So they say AI startup
(20:56):
perplexity has a valuation thatsurged to 14,000,000,000 in a
new funding round, and they'reraising 500,000,000. So what
this doesn't mean is that theVCs put in 14,000,000,000.
Now some sometimes companies doraise billions of dollars like
like OpenAI. But in this case,the company is raising
(21:19):
500,000,000. They're saying,we're gonna bring in
500,000,000. Those VC firms orwhoever it is put in that money
at a rate. So they buy a portionof the company with a
proportionality that would makethe company in theory worth
14,000,000,000, like thepercentage of the company that
(21:41):
they're buying for500,000,000,000.
Chris (21:43):
Multiply that out to a 100 percent Exactly. Of equity,
then yeah. Yes. Then that wouldbe the so it's it's a little bit
the the valuation is to somedegree, and and I remember this
from years and years ago doingfinance in graduate school, but
it's it's a little bitmanipulated in the sense of you
kind of come to a determinationof what percentage of the
(22:05):
company they're gonna get andwhat they're willing to put in
for that, and then it kind ofextrapolates out to what that
valuation is.
Daniel (22:10):
Yeah. Yeah. Exactly. And yeah. So this isn't a it's sort
of the the assumed, if you wannathink about it like the assumed
value, but it's the sort offair, maybe what you would call
the fair market value if you getan official valuation from the,
you know, that would stand up tothe IRS or something like that.
(22:31):
But this is kind of, yeah, it'show they kind of multiply
multiply that out. Yeah.
Chris (22:38):
Yeah. In theory, if you were to sell all the shares of
the company and everybody wasgonna buy in at that rate, at
that point in time, it wouldcome to that valuation.
Daniel (22:46):
Yeah. Now there's all sorts of trickiness that can
happen here because it also likeme as a founder, I I own a
certain number of shares ofPrediction Guard, right? And
other of our employees haveshares that have been granted to
them of Prediction Guard. Nowoften when you raise a round of
(23:10):
funding, in many cases, sharesmight be added to the overall
pool, right? So for example, theVC might say, well, we would
like to have this percentage orthis, you know, this many
shares, this is what we're gonnaget, but we also want you to
bump up the pool of shares thatyou have so that you can give
(23:33):
your future employee shares.
That sounds really great, but italso means that if you have more
shares now than right, my sharesare worth less diluted. So
that's what this sort ofdilution means if you've heard,
if you've heard that term. Yeah.Sure. Yeah.
That's all, I guess, the sort ofsetup, some of the terminology,
(23:58):
some of what happens in these. Iguess maybe the other
terminology to just use, whichis quite relevant here as we
kind of dig down into the sideof this, which is, you know, AI
startups. What's it what's itlike to raise what's it like to
go through this process ifyou're an AI startup? The the
other thing maybe just tohighlight before we get into
(24:19):
that are the what people oftencall the sort of stages of
funding. So you might have heardin the news, you know, this AI
startup raised Preseed, this oneraised Seed, this one raised
Series A, etcetera.
This is, you know, kind ofconfusing because this gives the
(24:41):
impression that these are verywell defined terms. They're
somewhat not well defined terms,but if generally you could think
about things as pre seed islike, have kind of an idea and
not really any traction, maybevery few, if any sales of your
product. Seed is maybe you doactually have a product. There's
(25:02):
some sales, some level oftraction, but certainly you
haven't proved out the fit ofyour product to the market. And
then series A is like, okay, wewe have repeatable sales.
We found a fit. Let's kind ofpour some gas on this, you know,
capital to actually get get moresales, get market share. And
(25:23):
Yeah. If you need to do thatmore, then you raise future,
know, series b, c, d, e, f, whatwhatever that that is. Right?
So for us in raising our roundthat that we just finished, we
raised a seed round. So I wouldsay in in some ways, we were
maybe in some ways, part of thedynamics of what I learned was
(25:46):
maybe we were a little bit pastwhere a lot of people called or
thought of as seed because wehad maybe more traction or sales
than that.
Chris (25:54):
Yeah. Had been bootstrapping a while there.
Daniel (25:56):
Yeah. We've been selling for a while. We had, you know, a
a non negligible amount ofcustomers and people were using
the product in production and wehad sales. And and so a certain
level of fit, but maybe notwhere you would say you would
wanna see at series a, where youjust it's a matter of cranking
the mechanism, right? We knowexactly where we're gonna get
(26:17):
all these customers, give us Xamount of dollars and you get X
amount of revenue out, right?
So in that sense, that's kind ofwhere we were. And one of the
dynamics that we found kind ofin the current market is there's
a lot of investors that investmaybe in seed or series A, like
(26:37):
both, right? They they may makeinvestments in either. Often
investors, specialize in certainstages like early, mid, late,
etcetera. So for us, when wewere talking to investors that
would invest in seed or seriesa, let's say, they look out at
the AI world and it's justabsolute pure chaos.
(27:00):
Right? So no one knows what thekiller AI app is. There's so
many AI startups. Everybody's anAI startup right now. No one
knows who's gonna win.
Right? And so I think for a lotof the companies that they're
looking at, right, they say,well, cool, prediction guard,
you're maybe what we wouldclassify as seed. You have great
(27:23):
traction, a really great team,which I think we do have, but
there's so much chaos in themarket and so much I am having
so much trouble differentiatingbetween all of these different
AI products. I'm just gonna waituntil series A. And if you make
it, like, cool.
Then then we'll let let's talkthen. Right? Because they're
(27:47):
very unsure about what's gonnakind of stand the test of time,
right, in a very kind of frothy,noisy AI market. So everything's
AI. It's really hard todifferentiate between kind of
different AI offerings just fromlooking at a pitch deck or even
having a half hour, an hourconversation.
(28:09):
You know, how really is LangFlowdifferent than LangChain,
different than LAMA index,different than XYZ, other
things, and how is thatdifferent from what you're doing
and how is that different fromwhat the hyperscalers are doing?
And and that sort of thing. So,it's that's one of the dynamics
(28:32):
that I was going to, going tohighlight is just this
frothiness, this noisiness inthe AI market in some ways. And
this is maybe not, it's a trendthat has happened with other
technology, but in some waysit's like, if you have a really
good team and you don't have,you haven't built your AI
(28:53):
product yet, then that's a greatplace for people to pour in
money because if it's a greatteam, you'll figure it out and
use our money well, maybe. Andif you have a product with
really solid fit and aconsistent kind of sales
process, then that's a greatplace to pour money in, right?
Because then like, it's just amatter of putting gas on the
(29:15):
fire. If you're in the middleand you have a product, right?
You put a flag in the ground inthis very noisy market, but it
hasn't really proved out yet interms of the fit. That's where,
you know, there's a lot ofuncertainty and the questions
are really about competition anddifferentiation.
Chris (29:33):
So as you've kind of personally gone through this
process, and, you know, I knowearlier on, we would be talking
between shows or and stuff, andyou'd be going around and and
kind of bootstrapping, you know,you were getting business and
you were getting everythingmoving and running a business in
those early days. And then thena little bit later, there was
(29:56):
the period where I'd we startedtalking and and you were you
were constantly dealing withfunding responsibilities as a as
a founder CEO. What surprisedyou is you're kind of going
through this learning process,and and you're you're trying to
accommodate that, and there wasclearly some sort of shift of
activities. And I'm not privy,you know, just I noticed that as
(30:18):
your friend, you know, that thatit seemed like you're always
doing these, and then it seemedfor a while you were doing
those. But, obviously, you stillhave to keep running the
business along the way.
How did that change theactivities and your
expectations, and yourperception of your own of your
own business while you'rerunning it?
Daniel (30:34):
Yeah. You know, it's it's it is really interesting to
go through this. So, you know,just by way of context, you
know, at least up until very,very recently, and I'm talking,
you know, in the the past monthor so, everyone in our company,
has been technical. Myself, youknow, our CTO, all of our
(30:55):
engineers, AI engineers, backend engineers, front end
engineers, we're just like, weknow how to build AI systems
well. That's our specialty andwe're all technical and that's
what we do.
And, and we're, you know, we'reproud of that. We're proud of
the kind of strong technicalbase. Right? And so as a result
(31:17):
of that, you know, I'm I'm theone I mean, partially because,
you know, we engage on thispodcast, you know, you and I
both engage in the AI communitydirectly. It's natural for me to
sort of shift into that.
Well, hey, if if there's apotential customer, I'm jumping
on a call with them to talkthrough whether this is a good
fit, do that discovery, youknow, eventually get through the
(31:41):
contracting, start apartnership, that sort of thing.
So I'm everyone in the companyis there we have no salesperson
up until recently. And, and so Iwas doing that piece. Well, if
that kind of shifts then totaking, you know, five to seven
VC pitches a day calls, thefollow-up, you know, paperwork
(32:06):
and like, hey, could you get meprojections for this and that?
Obviously, that sort of reducescapacity, right, for you to do
those other things.
Now, thankfully, like productwise, engineering wise, they're
rolling along. Right? Becauseour our CTO, you know, has it.
He could he can build out thatteam in his sleep and product
(32:27):
wise is good. But yeah, Idefinitely, you see that sort of
distraction on the sales side.
And part of what you're sayingis, part of why we're raising
this money is to kind of buildout more on that, go to market
side. I kinda just need to get Ineed to get this done and get
back to focus and get back tothis. Otherwise, you know, I
(32:48):
appreciate you all, but butwe're, like, time is ticking and
we've got a product that cansell and, you know, people are
eager and they see the need forthe product, but I just, like, I
can only have so manyconversations in a day. And if
those are five VC conversations,it's time that isn't with, you
know, five customers, right? So,that's really, you know, a thing
(33:13):
and something to definitelybalance if you're in those sort
of positions out there.
Yeah. It's something toconsider.
Chris (33:20):
I'm curious, you know, you touched on, you know, kind
of the notion of going frombeing a technologist and doing
the technology, you know, tokind of having to administer,
you know, the company as the CEOin that way. And, you know, how
hard is it? I think I myobservation was you've made this
transition very, very well. ButI am curious your own perception
(33:44):
because, know, for your in thoseearly stages, when you're when
you're just running the businessin that kind of bootstrap phase
that we talked about, you'rekind of doing you're doing sales
and stuff, but you're stillbuilding the company. And part
of that is it's a technologycompany.
And and I know darn well thatyou were coding and you were
doing all of the things, youknow, in those early days as it
(34:05):
was starting to grow and you'regoing in, you're transitioning
to this VC stage, how hardpersonally is it for you to let
go of those responsibilities anddistribute those out to people
on the team when you know maybein some cases, you know, I can
just knock this out. I'm reallygood at this particular skill. I
know how to do this thing. I'vedone it many times, but yet, you
(34:28):
know, you can't now. What's thatlike?
Daniel (34:30):
Well, and also a lot of those things I enjoy. Right?
It's like,
Chris (34:34):
I know you do.
Daniel (34:35):
You know? You you enjoy you know, we've talked about on
the show. I enjoy data munging.I enjoy, you know It's a sick
sick thing that
Chris (34:42):
you admit periodically. I know.
Daniel (34:45):
So, yeah, it definitely is a is a shift. I think, you
know, now in the in ourplatform, you know, which we're
so this is a self hosted AIsystem that we're deploying with
with customers. I don't knowthat any of my original code
exists anymore in that, in thatsystem. You know, I I built out
(35:08):
that first MVP along with ourkind of founding engineers.
Shout out to to Jake and Ed.
But that's been completelyrefactored at this point. And,
you know, I don't know if any ofthat survives. What I think I've
realized is a couple things. Oneis I personally love digging in
with customers and at theapplication level and really
(35:31):
more on the developer relationsside. That's why I do this
podcast.
Right? We're always talking moreto those practitioners, people
that are applying thetechnology. Right? That's a huge
passion of mine. And so that'sactually in our context, because
we're building this AI system,really the main work on the
(35:53):
product is really infrastructureand platform engineering and
that sort of thing.
There's obviously AI involved init because it's an AI system.
But really the AI stuff isreally in how people are using
the platform, right? Which is, Ithink seeing that as the
exciting piece to dig in was,was part of what happened. I
(36:14):
think the other thing that, wasactually something I think my,
my wife helped me see over timewas that you giving up, if you
wanna think about it as givingup you, you releasing the reins
on something that you're holdingtight to work wise, whether that
be a certain technology or apiece of the product that you're
(36:35):
digging into, or that sort ofthe thing, that sort of thing.
It's not, you don't have to viewit as a loss because basically
what you're doing is you're, isyou're creating an opportunity
that someone else can step intothat they didn't have before.
So one of the other engineers,one of the other team members,
as I delegate or release thosethings or step out of certain
(36:56):
responsibilities, it givessomeone else a challenge to step
up and an opportunity for themto learn and grow in their
career and actually be blessedimmensely to, have that
opportunity and to grow in thatway. And so I think as soon as I
started viewing that as a as anet good, not a, you know, a
(37:16):
positive, not a loss, that thisis actually creating some an
opportunity or an environmentwhere people can level up so
quickly as part of the joy ofstartups. Right? You get to do
all these things that you didn'tthink you'd get to do. That that
I think is a it helps my mind inin really understanding that.
And and so, yeah, it's awesometo see the team humming along
(37:37):
and the product coming comingalong. And it's also fun just to
go into our ticketing system andsay, Hey, what if we had this
feature and that and I can justcreate a bunch of tickets now,
and then I don't have toactually implement them. So it's
a lot easier to create ticketsnow.
Chris (37:53):
With, you know, we always wind up with what we kind of
internally call our our ideaabout the future, asking, you
know, the future when people aretalking. I wanna I wanna lead
into that for a moment and justsay, you know, you and I have
been doing the show now for, Iguess, it's been since mid two
the July 2018 when we firstreleased. So, you know, you're
(38:14):
you're talking, seven yearsnonstop of doing the show. And,
as as my friend, I have seen youevolve tremendously over that
time period. You you were verytechnically focused early on and
still are very technical,extremely technical.
But, you know, you've taken onthese new roles and evolved over
the years. And so the the theDan, who is my friend now, is
(38:38):
has evolved tremendously sincethe the the Dan that I first met
way back.
Daniel (38:42):
Thanks, man.
Chris (38:43):
I'm I'm curious. How do you see yourself in in like,
where do you see yourself goingin that? You know, kind of give
yourself your own futurequestion that you would be
asking the guest. Tiana, how doyou see things going forward as
we wind up?
Daniel (38:57):
Yeah. Well, I mean, of this is very much a more
philosophical answer, but I doreally view now, and I didn't
always see this, but, folks likethe, the Praxis Guild that I'm a
part of down in Indianapolis andthe work that they do have
helped me see that reallyventure building and those that
(39:21):
build ventures are venturebuilding is the thing that
shifts culture. Right? And sowhen I look out and I see all of
these AI companies that are forthe most part, very
exploitative, you know, notcalling out names here, but,
like, even what we started outthe conversation with, you know,
(39:43):
the fact that you wouldrepresent an AI company and just
have, you know, developers inIndia behind the scenes, like,
pretending to be bots. Likethings like that, you hear about
something like that, it seemslike every day.
And I think what I view asaspirational for us and what I
desire is to see our company seta different standard and
(40:04):
actually shift the culturearound actually enhancing or
advancing trust in humaninstitutions, in our case, via
private secure Gen AI systems.So that's really my vision for
what I would like to see there.And I think, it's viewing it
(40:25):
from that redemptive or thatrestorative perspective is
something that, that, yeah, isreally exciting for me. So,
yeah, I appreciate you going onthe journey with me, Chris. It's
been fun.
Chris (40:36):
Yeah, it's been it's been interesting to see you doing all
that over time. So as alwayswishing you all I think I'm
prediction guards biggest fanoutside the company. So thanks
for sharing today. This has beenreally fascinating for me. And
it's kind of just hearing thejourney.
I've heard snippets of it alongthe way, as you've talked, but
(40:56):
actually, we've never done this,you know, deep dive into the
conversation. So thanks forsharing with me and the
audience.
Daniel (41:01):
Yeah, of course, Chris. We'll talk to you soon. Take
care, man.
Jerod (41:10):
Alright. That's our show for this week. If you haven't
checked out our website, head topracticalai.fm and be sure to
connect with us on LinkedIn, X,or Blue Sky. You'll see us
posting insights related to thelatest AI developments, and we
would love for you to join theconversation. Thanks to our
partner Prediction Guard forproviding operational support
for the show.
(41:30):
Check them out atpredictionguard.com. Also,
thanks to Breakmaster Cylinderfor the beats and to you for
listening. That's all for now,but you'll hear from us again
next week.
Welcome to the Practical AI podcast, where we break down
the real world applications ofartificial intelligence and how
it's shaping the way we live,work, and create. Our goal is to
help make AI technologypractical, productive, and
accessible to everyone. Whetheryou're a developer, business
leader, or just curious aboutthe tech behind the buzz, you're
(00:24):
in the right place. Be sure toconnect with us on LinkedIn, X,
or Blue Sky to stay up to datewith episode drops, behind the
scenes content, and AI insights.You can learn more at
practicalai.fm.
Now, onto the show.
Daniel (00:48):
Welcome to another fully connected episode of the
Practical AI Podcast. In thesefully connected episodes, it's
just Chris and I, and we spendsome time catching up on AI news
and and maybe some other topicsthat are just on our mind and
hopefully share some things withyou that will be useful and
(01:09):
educational as you explore theAI world and and your own, your
own work with AI. I'm DanielWitenack. I am CEO at Prediction
Guard, and I am joined as alwaysby my cohost, Chris Benson, who
is a principal AI researchengineer at Lockheed Martin. How
are doing, Chris?
Chris (01:28):
Doing well today, Daniel. So much is happening. Looking
forward to our conversationtoday. Yeah. So much is
happening.
Daniel (01:35):
It's summer is upon us, which is is great. If my
allergies are any indication,then then things are things are
changing. But, yeah, there's athere's a lot of interesting
things changing and shiftshappening in in the AI world.
One of the things that I'm I'mconstantly just I see all the
(01:55):
time in the news, we we have a ateam, you know, a channel in our
in our Slack, just a randomchannel where people share news
stories. And there's alwayssomething about, you know, AI
startups, either good or bad ofof AI startups.
Actually, this this last week,one of the one of the things
(02:16):
that was shared was this, youknow, hyped AI startup that that
raised a bunch of money, wasvalued at 1,500,000,000.0,
Builder dot ai or builder.ai.And I think, basically, it it
collapsed. There were sort ofthese these I I don't know
everything about it, but I sawsort of posts about the AI not
(02:41):
being AI, but actually being,you know, Indian coders or
developers. So it was like a abuilder platform, like a low
code, you know, I'm gonna buildan application, but actually,
there's some indication thatthere was actually just people
building building behind thebackground. I I don't know the
full story behind that or ifthat's even true.
(03:01):
But regardless of the sort ofcollapse, I don't know if you're
seeing a similar flurry of allsorts of discussions of AI
companies, good or bad.
Chris (03:11):
Yeah. And and, you know, there's there's kind of never
ending talk of AI bubbles and,you know, fundings up and down.
And and I know you as a as astartup founder yourself have
have a lot of interesting,experiences with the funding
world and stuff like that. Maybemaybe get you to share some of
(03:31):
that today.
Daniel (03:32):
Yeah. Yeah. Sure. A little bit of the what what is
it what is it actually like inthe trenches, to raise funding
for for an AI startup? What whatperceptions are there?
What questions come up? Youknow, what stages are there? I'm
I'm just looking at I search forAI funding startup in in Google
(03:54):
News, and it's just sort of a aconstant feed. The ones that are
popping up as we are recordingthis episode, legal startup
Harvey AI raising funding at a$5,000,000,000 valuation,
Decagon to raise a 100,000,000at 1,500,000,000.0 valuation,
Samaya AI, gets 43,000,000.Perplexity is raising around
(04:21):
bringing in 500,000,000 at a14,000,000,000 valuation.
Yeah. It may be even useful forsome people out there that
aren't as much in the startupworld to even just parse through
some of what this actually eveneven means, like some of these
(04:42):
terms. So maybe that would be agood set of first questions to
answer is what does it actuallymean to raise funding regardless
of whether you're an AI startupor not?
Chris (04:53):
And maybe even before we, you know, just as a part of
that, even what some of thatbasic, terminology means, you
know, where you're where you'reraising a certain amount at a
certain valuation. Yeah. Forthose who have not been exposed
to to the finance side of thingsin the past, you know, and and
what is the how does the thelevel of valuation affect your
funding and affects control,things like that? So
Daniel (05:16):
Yeah. Yeah. Well, there's I I have to admit
myself, you know, learning a lotabout this in the in the past
year as as we've gone throughthis process. Lots of
interesting, strange, you know,terms and and processes and and
all of that. It's a whole worldin and of itself.
But yeah, so just if you thinkabout building a company, you
(05:40):
know, you're gonna build an AIcompany. Typically that company
might be fall into a fewdifferent categories, right? So
maybe you're an AI company thatyou want to provide services to
others. Like you wanna help thembuild AI things. This would be
like a service provider type ofcompany, a consultancy, that
(06:05):
sort of thing.
Typically, when you're hearingabout these companies raising
venture capital, which we'lltalk about that here in a
second, they're not reallyfocused on these sort of service
provider companies. There isactually a huge market for that
right now. Actually, I think, Iwas on a call with someone and
it's like, you know, OpenAIisn't the one really making the
(06:27):
the most money off of of AI.It's like Accenture. Like, the
the consultancies of the worldare are just making a a huge
amount of money off of AIbecause people, you know, don't
know how to how to build thesethings.
They they don't know how toadopt things, etcetera. So it is
a really good space to be in.But in in another category, your
(06:48):
AI startup might be a a productbased company. So you have a,
you know, an AI agent platformthat you're building for
petroleum engineering, or youhave a healthcare assistant
company that you're building, oryou have a developer tools
company that has developer toolsfor working with AI models or
(07:09):
platform or infrastructuresoftware. In our case for
Prediction Guard, we're more onthat infrastructure system side,
right?
But all of these are kind ofproducts. Now, if you then kind
of subdivide those products,there's a set of products that
are probably somewhat niche.Like you have an AI product, but
(07:35):
the number of people that aregonna use it, if you wanna think
about it, the market for that AIproduct is fairly small, right?
So maybe I have, I don't know,I'm thinking off the top of my
head, but I have a an AI tool oran AI product for, pug owners. I
(07:57):
had a I had a pug dog for awhile.
So I I have an AI product forpug owners, and it helps, you
know, answer questions aboutshedding or about great things
to do with your pug. Now maybe II don't know how many pug owners
there are in the world, butmaybe that's not, like, the the
biggest market that you couldthink of in terms of the amount
(08:18):
of money you could make off ofsuch a thing. I don't know if
you agree or not, Chris.
Chris (08:22):
Oh, I don't know. There's lots of pug owners out there.
Daniel (08:25):
Very true. Very true. There may be those products and
you could kind of think aboutthose if you want, like some
people might call these sort oflifestyle type of businesses
where you could actually, youcould probably make a reasonable
amount of money. Now, I don'tknow if the Pug AI would make a
(08:47):
reasonable amount of money, butyou could make a good living for
yourself, maybe.
Chris (08:52):
It it would be. So I think kind of to the point,
there's an addressable marketthat that you would that your
startup would need to be wouldneed to be working towards.
Daniel (09:03):
Correct. And
Chris (09:04):
it needs to be an addressable market that makes
what you're doing worthwhile.
Daniel (09:07):
Correct. Yes. And and sometimes there's sort of two
there's really two things youcould determine once you look at
the market that is available forwhat you're building. On one
side, there could be a marketand that market could be big
enough to support your business,maybe even employees pay you a
paycheck. Right?
(09:28):
But it's not, it's never goingto be a large market in the
sense of billions of dollars. Soreally, you know, billions or
tens of billions or hundreds ofbillions of dollars. Right? Then
there's another set of productsthat would have a larger market.
So, you know, just by way ofexample, you look at certain
(09:52):
startups from the past thathave, you know, captured a very
large market, whether that be,you know, Uber or Airbnb, I
guess those are consumer relatedthings.
But then, you know, there's allsorts of things in the computing
and B2B side. So you look atsomething like Docker or, you
(10:14):
know, whatever these kind oflarger, there's a huge market
for some of these types ofproducts. So if your AI product
doesn't just have a market of,you know, some millions of
dollars that could support yourlifestyle, but it's, you know,
tens of billions of dollars orhundreds of billions of dollars
(10:34):
of market out there, then thatkind of categorizes you in the
potential spot where you couldraise venture capital for your
startup. And so now we kind ofneed to define what that means,
which is if you're gonna raiseventure capital, basically
(10:55):
venture capitalists arecompanies that they actually
have their own investors. Sothey get together a bunch of
investors often called LPs orlimited partners that put in
money into a fund.
Let's say it's a $50,000,000fund and the VC firm then
(11:18):
allocates from that fund toinvest in a number of companies,
in quite a few companiesactually, with the hopes that
certain of these companiesactually do scale up to reach a
reasonable amount of that verylarge market, thus becoming
(11:40):
large valuable companies thatprovide a return to that
investor investor pool.
Chris (11:49):
I'm curious. One of the often before the VC stage,
you'll have angel investors,which are, you know, wealthy
individuals that are getting invery, very early Yep. Even
before the VC do and do that. Doyou have did did you guys go
that way? Or or or at least areare you seeing that in the AI
space?
Daniel (12:07):
We we didn't. In in our case, actually, we spent really
a year and a half, the firstyear to year and a half of our
business bootstrapping thebusiness, which what that that
means is basically, you know,businesses ideally make money.
And if you get an, know, if yousell enough of what you're
making and you bring money in,you can pay bills and keep
(12:31):
running your business. Right?And so that's kind of, you're
living essentially growing yourbusiness as fast as it can based
on the money that's coming inrevenue wise.
And so we did that initially.Others kind of might not be in
the place where they do that.They might not be able to get
contracts or sales early enough.And so maybe there's kind of
(12:52):
some initial money in from someangels.
Chris (12:55):
I've got a question for you on that. And that is,
there's a point there, you know,where you're bootstrapping early
on. And I remember when you weregoing through that process, and
there was a point, you know,along the way, and I don't know
where that happened, but whereyou decide you are in, you need
to go the VC route to achieveyour goals and stuff. How what's
your thinking as an AI founderabout that? You know?
(13:19):
Because Yeah. There are tradeoffs on on if you bootstrap, you
know, all the way through andgrow based on your own revenues,
you know, that gives you a acertain advantage in some ways
that, you know, you're nothaving to go through the process
of VC. As a founder, how did youhow did you really kind of
decide, you know, what the tradeoffs were and, you know, and and
(13:40):
go the way that you ended upgoing?
Daniel (13:42):
Yeah. It's a good question, and I don't think
there's any right answer to thisquestion for different people. I
think it's a question peoplehave to answer for themselves
and the business that they'rebuilding. For us, I think what
we saw was that we did have aproduct that we were creating,
(14:03):
which at least had some level offit to a real need in the
industry. And there was thechance to scale that product up
and gain market share, but onlyif we did that sort of quite
rapidly.
Because a lot of theconsolidation, I think the kind
(14:25):
of acquisition and consolidationthat will happen in the AI
market, it's not gonna happen,you know, ten or fifteen years
from now. It's gonna happen, youknow, three to five years or may
or maybe sooner from now. Right?So if you want to kind of
participate in that sort ofconsolidation, potentially be
acquired, then you really needto to kind of, scale that quite
(14:48):
quickly, which requires capital.It's not that you couldn't grow
organically, right, and stillgrow organically and have an
again, a nice lifestylebusiness, but you probably
wouldn't participate in thatkind of consolidation phase.
For us, that's ultimately whatwe decided. Is risky, but that's
(15:08):
ultimately what we decided todo.
Sponsor (15:24):
You know, with everything going on in the AI
world right now, it is easy tofeel overwhelmed. There's a lot
of noise out there. People arewondering if AI is coming for
their jobs. But Miro's latestsurvey paints a more optimistic
picture. Here's what it says.
76% of knowledge workersactually believe AI can help
their roles, not take them. Butthe problem is, over half those
(15:48):
folks still don't know when orhow to use AI. That's where
Miro's Innovation Workspacecomes into play and has become a
go to for our team. We've triedit out. We love it.
Whether we're planning anepisode, building out a new
podcast format for our client orjust brainstorming ideas, What
used to take hours across docs,whiteboards, chat threads. Now
(16:09):
it all happens in one space. Andhonestly, it's a lot of fun.
Miro makes collaborating withyour team and with your A. I.
Way less intimidating and waymore productive. The coolest
part is you don't have to be anexpert with A. I. The work
you're already doing with Miro'sCanvas becomes the prompt. You
can turn a sticky note intodiagrams, screenshots into data
(16:31):
tables and ideas into realoutcomes fast.
So help your team get great donewith Miro. Check out miro.com to
find out how that is mir0.com,miro.com and tell them you heard
about it on Practical AI.
Daniel (16:54):
Well, Chris, we talked we talked a little bit about
what we were trying to do inthis fundraising phase. We
defined a few terms like venturecapitalists. I think it might be
interesting for people to knowmaybe that haven't been exposed
to this, some of the dynamicsthat venture capitalists might
might be thinking of. Now I feelreally privileged that, that
(17:17):
that we've got some folks thathave partnered with with
Prediction Guard that arereally, with us and supporting
us in in a really, profound way,also, you know, aligning to a
lot of how we wanna build thebusiness. But if you just kind
of strip away anything but,like, the capital and return
pieces of how at least many VCsmight think, Again, going back
(17:42):
to the example that we hadbefore, there might be, you
might have a fund of 50,000,000or a $100,000,000 and you invest
in however many, let's say 30different companies.
What you're hoping for is, andyou make an investment in that
company. Essentially what you'redoing is you're, you're, the
founders are selling a piece oftheir company and you're buying
(18:04):
a piece of that company viaequity. So you're giving away
shares or equity in yourcompany, a percentage of your
company. There's various typesof mechanisms of how that
happened, whether it's just, youknow, you value the company and
you buy certain shares or youbuy the right to, you know, have
a certain percentage of thecompany later on that converts.
(18:25):
There's all sorts of differentmechanisms around how that
happens, safes and convertiblenotes and venture debt and
priced, valuation rounds, thatsort of thing.
But essentially that's whatyou're doing. And what the VC
would hope for, let's say thatwe do that with 30, you know,
Chris, you and I have$50,000,000, which I don't think
(18:48):
we do. But let's just say wedid. We don't. And we we put you
know, we divide that out across30 different companies.
We buy pieces of thesecompanies. Essentially, this is
a very risky thing to do, butwhat you're hoping is that it's
sort of like a power laws sortof thing. So you're kind of
(19:08):
hoping that one of thosecompanies just is awesome and
grows like wildfire, is acquiredfor a huge amount of money, and
that your one investment in thatone company returns you your
whole fund. $50,000,000, let'ssay. Then another three of those
(19:28):
returns your fund again.
So not as good of exits, right,but, you know, not bad. So you
basically have doubled yourmoney. And then the whole
remaining set of the companiesreturns it maybe a third time.
Right? So essentially you'vetripled your money, but most of
that has come from a very smallamount of the companies that
(19:50):
you've invested in, and you geta good, you know, you get a good
return for your investors whoput the money into to your fund.
Chris (19:58):
It's a form of diversification of risk.
Daniel (20:00):
Yeah. Essentially.
Chris (20:00):
So you're you're selecting a a bunch of different
companies and hoping that one ora a small number of them will
will far outpace Yeah. The thelarger collective.
Daniel (20:11):
Yeah. Now that's not the rule across the board, but it's
kind of generally how you couldthink of this. There's different
firms that have differentstrategies that make a smaller
number of investments or maybemake a very, very, very large
number of very smallinvestments. Right? And so
there's there's differentstrategies within that, and
that's kind of, part of that VCworld and and determining what
(20:35):
the, you know, the way that youwanna run your fund and the
thesis that you have and all ofthose sorts of things.
But that's kinda what'shappening. So if we just go back
to kind of interpreting some ofthese things. So I mentioned
that, perplexity raise, right?So they say AI startup
(20:56):
perplexity has a valuation thatsurged to 14,000,000,000 in a
new funding round, and they'reraising 500,000,000. So what
this doesn't mean is that theVCs put in 14,000,000,000.
Now some sometimes companies doraise billions of dollars like
like OpenAI. But in this case,the company is raising
(21:19):
500,000,000. They're saying,we're gonna bring in
500,000,000. Those VC firms orwhoever it is put in that money
at a rate. So they buy a portionof the company with a
proportionality that would makethe company in theory worth
14,000,000,000, like thepercentage of the company that
(21:41):
they're buying for500,000,000,000.
Chris (21:43):
Multiply that out to a 100 percent Exactly. Of equity,
then yeah. Yes. Then that wouldbe the so it's it's a little bit
the the valuation is to somedegree, and and I remember this
from years and years ago doingfinance in graduate school, but
it's it's a little bitmanipulated in the sense of you
kind of come to a determinationof what percentage of the
(22:05):
company they're gonna get andwhat they're willing to put in
for that, and then it kind ofextrapolates out to what that
valuation is.
Daniel (22:10):
Yeah. Yeah. Exactly. And yeah. So this isn't a it's sort
of the the assumed, if you wannathink about it like the assumed
value, but it's the sort offair, maybe what you would call
the fair market value if you getan official valuation from the,
you know, that would stand up tothe IRS or something like that.
(22:31):
But this is kind of, yeah, it'show they kind of multiply
multiply that out. Yeah.
Chris (22:38):
Yeah. In theory, if you were to sell all the shares of
the company and everybody wasgonna buy in at that rate, at
that point in time, it wouldcome to that valuation.
Daniel (22:46):
Yeah. Now there's all sorts of trickiness that can
happen here because it also likeme as a founder, I I own a
certain number of shares ofPrediction Guard, right? And
other of our employees haveshares that have been granted to
them of Prediction Guard. Nowoften when you raise a round of
(23:10):
funding, in many cases, sharesmight be added to the overall
pool, right? So for example, theVC might say, well, we would
like to have this percentage orthis, you know, this many
shares, this is what we're gonnaget, but we also want you to
bump up the pool of shares thatyou have so that you can give
(23:33):
your future employee shares.
That sounds really great, but italso means that if you have more
shares now than right, my sharesare worth less diluted. So
that's what this sort ofdilution means if you've heard,
if you've heard that term. Yeah.Sure. Yeah.
That's all, I guess, the sort ofsetup, some of the terminology,
(23:58):
some of what happens in these. Iguess maybe the other
terminology to just use, whichis quite relevant here as we
kind of dig down into the sideof this, which is, you know, AI
startups. What's it what's itlike to raise what's it like to
go through this process ifyou're an AI startup? The the
other thing maybe just tohighlight before we get into
(24:19):
that are the what people oftencall the sort of stages of
funding. So you might have heardin the news, you know, this AI
startup raised Preseed, this oneraised Seed, this one raised
Series A, etcetera.
This is, you know, kind ofconfusing because this gives the
(24:41):
impression that these are verywell defined terms. They're
somewhat not well defined terms,but if generally you could think
about things as pre seed islike, have kind of an idea and
not really any traction, maybevery few, if any sales of your
product. Seed is maybe you doactually have a product. There's
(25:02):
some sales, some level oftraction, but certainly you
haven't proved out the fit ofyour product to the market. And
then series A is like, okay, wewe have repeatable sales.
We found a fit. Let's kind ofpour some gas on this, you know,
capital to actually get get moresales, get market share. And
(25:23):
Yeah. If you need to do thatmore, then you raise future,
know, series b, c, d, e, f, whatwhatever that that is. Right?
So for us in raising our roundthat that we just finished, we
raised a seed round. So I wouldsay in in some ways, we were
maybe in some ways, part of thedynamics of what I learned was
(25:46):
maybe we were a little bit pastwhere a lot of people called or
thought of as seed because wehad maybe more traction or sales
than that.
Chris (25:54):
Yeah. Had been bootstrapping a while there.
Daniel (25:56):
Yeah. We've been selling for a while. We had, you know, a
a non negligible amount ofcustomers and people were using
the product in production and wehad sales. And and so a certain
level of fit, but maybe notwhere you would say you would
wanna see at series a, where youjust it's a matter of cranking
the mechanism, right? We knowexactly where we're gonna get
(26:17):
all these customers, give us Xamount of dollars and you get X
amount of revenue out, right?
So in that sense, that's kind ofwhere we were. And one of the
dynamics that we found kind ofin the current market is there's
a lot of investors that investmaybe in seed or series A, like
(26:37):
both, right? They they may makeinvestments in either. Often
investors, specialize in certainstages like early, mid, late,
etcetera. So for us, when wewere talking to investors that
would invest in seed or seriesa, let's say, they look out at
the AI world and it's justabsolute pure chaos.
(27:00):
Right? So no one knows what thekiller AI app is. There's so
many AI startups. Everybody's anAI startup right now. No one
knows who's gonna win.
Right? And so I think for a lotof the companies that they're
looking at, right, they say,well, cool, prediction guard,
you're maybe what we wouldclassify as seed. You have great
(27:23):
traction, a really great team,which I think we do have, but
there's so much chaos in themarket and so much I am having
so much trouble differentiatingbetween all of these different
AI products. I'm just gonna waituntil series A. And if you make
it, like, cool.
Then then we'll let let's talkthen. Right? Because they're
(27:47):
very unsure about what's gonnakind of stand the test of time,
right, in a very kind of frothy,noisy AI market. So everything's
AI. It's really hard todifferentiate between kind of
different AI offerings just fromlooking at a pitch deck or even
having a half hour, an hourconversation.
(28:09):
You know, how really is LangFlowdifferent than LangChain,
different than LAMA index,different than XYZ, other
things, and how is thatdifferent from what you're doing
and how is that different fromwhat the hyperscalers are doing?
And and that sort of thing. So,it's that's one of the dynamics
(28:32):
that I was going to, going tohighlight is just this
frothiness, this noisiness inthe AI market in some ways. And
this is maybe not, it's a trendthat has happened with other
technology, but in some waysit's like, if you have a really
good team and you don't have,you haven't built your AI
(28:53):
product yet, then that's a greatplace for people to pour in
money because if it's a greatteam, you'll figure it out and
use our money well, maybe. Andif you have a product with
really solid fit and aconsistent kind of sales
process, then that's a greatplace to pour money in, right?
Because then like, it's just amatter of putting gas on the
(29:15):
fire. If you're in the middleand you have a product, right?
You put a flag in the ground inthis very noisy market, but it
hasn't really proved out yet interms of the fit. That's where,
you know, there's a lot ofuncertainty and the questions
are really about competition anddifferentiation.
Chris (29:33):
So as you've kind of personally gone through this
process, and, you know, I knowearlier on, we would be talking
between shows or and stuff, andyou'd be going around and and
kind of bootstrapping, you know,you were getting business and
you were getting everythingmoving and running a business in
those early days. And then thena little bit later, there was
(29:56):
the period where I'd we startedtalking and and you were you
were constantly dealing withfunding responsibilities as a as
a founder CEO. What surprisedyou is you're kind of going
through this learning process,and and you're you're trying to
accommodate that, and there wasclearly some sort of shift of
activities. And I'm not privy,you know, just I noticed that as
(30:18):
your friend, you know, that thatit seemed like you're always
doing these, and then it seemedfor a while you were doing
those. But, obviously, you stillhave to keep running the
business along the way.
How did that change theactivities and your
expectations, and yourperception of your own of your
own business while you'rerunning it?
Daniel (30:34):
Yeah. You know, it's it's it is really interesting to
go through this. So, you know,just by way of context, you
know, at least up until very,very recently, and I'm talking,
you know, in the the past monthor so, everyone in our company,
has been technical. Myself, youknow, our CTO, all of our
(30:55):
engineers, AI engineers, backend engineers, front end
engineers, we're just like, weknow how to build AI systems
well. That's our specialty andwe're all technical and that's
what we do.
And, and we're, you know, we'reproud of that. We're proud of
the kind of strong technicalbase. Right? And so as a result
(31:17):
of that, you know, I'm I'm theone I mean, partially because,
you know, we engage on thispodcast, you know, you and I
both engage in the AI communitydirectly. It's natural for me to
sort of shift into that.
Well, hey, if if there's apotential customer, I'm jumping
on a call with them to talkthrough whether this is a good
fit, do that discovery, youknow, eventually get through the
(31:41):
contracting, start apartnership, that sort of thing.
So I'm everyone in the companyis there we have no salesperson
up until recently. And, and so Iwas doing that piece. Well, if
that kind of shifts then totaking, you know, five to seven
VC pitches a day calls, thefollow-up, you know, paperwork
(32:06):
and like, hey, could you get meprojections for this and that?
Obviously, that sort of reducescapacity, right, for you to do
those other things.
Now, thankfully, like productwise, engineering wise, they're
rolling along. Right? Becauseour our CTO, you know, has it.
He could he can build out thatteam in his sleep and product
(32:27):
wise is good. But yeah, Idefinitely, you see that sort of
distraction on the sales side.
And part of what you're sayingis, part of why we're raising
this money is to kind of buildout more on that, go to market
side. I kinda just need to get Ineed to get this done and get
back to focus and get back tothis. Otherwise, you know, I
(32:48):
appreciate you all, but butwe're, like, time is ticking and
we've got a product that cansell and, you know, people are
eager and they see the need forthe product, but I just, like, I
can only have so manyconversations in a day. And if
those are five VC conversations,it's time that isn't with, you
know, five customers, right? So,that's really, you know, a thing
(33:13):
and something to definitelybalance if you're in those sort
of positions out there.
Yeah. It's something toconsider.
Chris (33:20):
I'm curious, you know, you touched on, you know, kind
of the notion of going frombeing a technologist and doing
the technology, you know, tokind of having to administer,
you know, the company as the CEOin that way. And, you know, how
hard is it? I think I myobservation was you've made this
transition very, very well. ButI am curious your own perception
(33:44):
because, know, for your in thoseearly stages, when you're when
you're just running the businessin that kind of bootstrap phase
that we talked about, you'rekind of doing you're doing sales
and stuff, but you're stillbuilding the company. And part
of that is it's a technologycompany.
And and I know darn well thatyou were coding and you were
doing all of the things, youknow, in those early days as it
(34:05):
was starting to grow and you'regoing in, you're transitioning
to this VC stage, how hardpersonally is it for you to let
go of those responsibilities anddistribute those out to people
on the team when you know maybein some cases, you know, I can
just knock this out. I'm reallygood at this particular skill. I
know how to do this thing. I'vedone it many times, but yet, you
(34:28):
know, you can't now. What's thatlike?
Daniel (34:30):
Well, and also a lot of those things I enjoy. Right?
It's like,
Chris (34:34):
I know you do.
Daniel (34:35):
You know? You you enjoy you know, we've talked about on
the show. I enjoy data munging.I enjoy, you know It's a sick
sick thing that
Chris (34:42):
you admit periodically. I know.
Daniel (34:45):
So, yeah, it definitely is a is a shift. I think, you
know, now in the in ourplatform, you know, which we're
so this is a self hosted AIsystem that we're deploying with
with customers. I don't knowthat any of my original code
exists anymore in that, in thatsystem. You know, I I built out
(35:08):
that first MVP along with ourkind of founding engineers.
Shout out to to Jake and Ed.
But that's been completelyrefactored at this point. And,
you know, I don't know if any ofthat survives. What I think I've
realized is a couple things. Oneis I personally love digging in
with customers and at theapplication level and really
(35:31):
more on the developer relationsside. That's why I do this
podcast.
Right? We're always talking moreto those practitioners, people
that are applying thetechnology. Right? That's a huge
passion of mine. And so that'sactually in our context, because
we're building this AI system,really the main work on the
(35:53):
product is really infrastructureand platform engineering and
that sort of thing.
There's obviously AI involved init because it's an AI system.
But really the AI stuff isreally in how people are using
the platform, right? Which is, Ithink seeing that as the
exciting piece to dig in was,was part of what happened. I
(36:14):
think the other thing that, wasactually something I think my,
my wife helped me see over timewas that you giving up, if you
wanna think about it as givingup you, you releasing the reins
on something that you're holdingtight to work wise, whether that
be a certain technology or apiece of the product that you're
(36:35):
digging into, or that sort ofthe thing, that sort of thing.
It's not, you don't have to viewit as a loss because basically
what you're doing is you're, isyou're creating an opportunity
that someone else can step intothat they didn't have before.
So one of the other engineers,one of the other team members,
as I delegate or release thosethings or step out of certain
(36:56):
responsibilities, it givessomeone else a challenge to step
up and an opportunity for themto learn and grow in their
career and actually be blessedimmensely to, have that
opportunity and to grow in thatway. And so I think as soon as I
started viewing that as a as anet good, not a, you know, a
(37:16):
positive, not a loss, that thisis actually creating some an
opportunity or an environmentwhere people can level up so
quickly as part of the joy ofstartups. Right? You get to do
all these things that you didn'tthink you'd get to do. That that
I think is a it helps my mind inin really understanding that.
And and so, yeah, it's awesometo see the team humming along
(37:37):
and the product coming comingalong. And it's also fun just to
go into our ticketing system andsay, Hey, what if we had this
feature and that and I can justcreate a bunch of tickets now,
and then I don't have toactually implement them. So it's
a lot easier to create ticketsnow.
Chris (37:53):
With, you know, we always wind up with what we kind of
internally call our our ideaabout the future, asking, you
know, the future when people aretalking. I wanna I wanna lead
into that for a moment and justsay, you know, you and I have
been doing the show now for, Iguess, it's been since mid two
the July 2018 when we firstreleased. So, you know, you're
(38:14):
you're talking, seven yearsnonstop of doing the show. And,
as as my friend, I have seen youevolve tremendously over that
time period. You you were verytechnically focused early on and
still are very technical,extremely technical.
But, you know, you've taken onthese new roles and evolved over
the years. And so the the theDan, who is my friend now, is
(38:38):
has evolved tremendously sincethe the the Dan that I first met
way back.
Daniel (38:42):
Thanks, man.
Chris (38:43):
I'm I'm curious. How do you see yourself in in like,
where do you see yourself goingin that? You know, kind of give
yourself your own futurequestion that you would be
asking the guest. Tiana, how doyou see things going forward as
we wind up?
Daniel (38:57):
Yeah. Well, I mean, of this is very much a more
philosophical answer, but I doreally view now, and I didn't
always see this, but, folks likethe, the Praxis Guild that I'm a
part of down in Indianapolis andthe work that they do have
helped me see that reallyventure building and those that
(39:21):
build ventures are venturebuilding is the thing that
shifts culture. Right? And sowhen I look out and I see all of
these AI companies that are forthe most part, very
exploitative, you know, notcalling out names here, but,
like, even what we started outthe conversation with, you know,
(39:43):
the fact that you wouldrepresent an AI company and just
have, you know, developers inIndia behind the scenes, like,
pretending to be bots. Likethings like that, you hear about
something like that, it seemslike every day.
And I think what I view asaspirational for us and what I
desire is to see our company seta different standard and
(40:04):
actually shift the culturearound actually enhancing or
advancing trust in humaninstitutions, in our case, via
private secure Gen AI systems.So that's really my vision for
what I would like to see there.And I think, it's viewing it
(40:25):
from that redemptive or thatrestorative perspective is
something that, that, yeah, isreally exciting for me. So,
yeah, I appreciate you going onthe journey with me, Chris. It's
been fun.
Chris (40:36):
Yeah, it's been it's been interesting to see you doing all
that over time. So as alwayswishing you all I think I'm
prediction guards biggest fanoutside the company. So thanks
for sharing today. This has beenreally fascinating for me. And
it's kind of just hearing thejourney.
I've heard snippets of it alongthe way, as you've talked, but
(40:56):
actually, we've never done this,you know, deep dive into the
conversation. So thanks forsharing with me and the
audience.
Daniel (41:01):
Yeah, of course, Chris. We'll talk to you soon. Take
care, man.
Jerod (41:10):
Alright. That's our show for this week. If you haven't
checked out our website, head topracticalai.fm and be sure to
connect with us on LinkedIn, X,or Blue Sky. You'll see us
posting insights related to thelatest AI developments, and we
would love for you to join theconversation. Thanks to our
partner Prediction Guard forproviding operational support
for the show.
(41:30):
Check them out atpredictionguard.com. Also,
thanks to Breakmaster Cylinderfor the beats and to you for
listening. That's all for now,but you'll hear from us again
next week.
Popular Podcasts
Cold Case Files: Miami
Joyce Sapp, 76; Bryan Herrera, 16; and Laurance Webb, 32—three Miami residents whose lives were stolen in brutal, unsolved homicides. Cold Case Files: Miami follows award‑winning radio host and City of Miami Police reserve officer Enrique Santos as he partners with the department’s Cold Case Homicide Unit, determined family members, and the advocates who spend their lives fighting for justice for the victims who can no longer fight for themselves.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.