Justin McCallon, founder and CEO of Callidus AI, discusses transitioning from corporate America to startup life. He shares insights on building versatile teams and the role of AI in modern law.
• Transitioning from corporate roles to entrepreneurship
• Importance of visionary thinking in startups
• Cross-functional skill sets as a necessity for startup success
• Challenges of AI adoption in the legal profession
• Future predictions for AI integration in law
• Advice for corporate professionals considering startups
All Links: linktr.ee/startup_recruiting
LinkedIn: www.linkedin.com/in/riecekeck/
Twitter/X: x.com/tech_headhunter
Recruitment: www.mindhire.ai
Youtube: https://www.youtube.com/@seedtoexitpod
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
When I try to think about vision and try to ask them
tough vision related questionsabout how could we make an
impact on the space, what I'venoticed is, especially in
corporate America, people thinkpretty small relative to in
startups.
When I talk to other founders,they're thinking really big, and
if people are thinking abouthow is the landscape going to
(00:21):
change and how they're going tocapture value from that and grow
their business in creative ways, that's a really good sign.
Definitely trying to get asense of can the person wear
multiple hats?
Everybody in a startup has to,even my developers.
Even though they're reallytechnical, they need to quickly
understand the legalimplications for what they're
(00:42):
building, and so we need peoplethat are cross-functional, um,
at least in the way that we dobusiness, and so I'd look for
people with those kind ofcross-functional abilities and
just high levels of curiosityand interest and drive.
Speaker 2 (00:57):
Thanks for listening.
To Seed to Exit.
I'm Rhys Keck and today I'mjoined by Justin McCallan,
founder and CEO of Calidus AI.
Calidus is an AI companydedicated to augmenting not
replacing, lawyers throughadvanced AI solutions.
Prior to founding Calidus,justin practiced law as a
commercial restructuring and M&Aattorney.
He then spent 12 years at AT&Tand its subsidiaries, including
Zander and DirecTV, includingwhere he launched the first
(01:20):
generative AI product forDirecTV.
In our conversation, we'regoing to talk about the
inspiration for Calidus, what itwas like to transition from a
large enterprise to a startup,the implications of AI in the
legal profession, fundraisinginsights and some advice for
inspiring entrepreneurs.
Really excited for you tolisten.
Let's go ahead and dive intothe episode.
Welcome to Seed to Exit, thepodcast where we uncover the
(01:42):
stories, strategies and insightsthat power the startup
ecosystem.
I'm your host, rhys Keck,founder of MindHire, a talent
acquisition firm specializing inhelping startups build
exceptional teams.
Each week, I sit down withfounders, investors and industry
leaders to explore the journeysbehind iconic companies and
game-changing ideas.
Whether you're building,investing or just curious about
(02:04):
what it takes to succeed in thestartup world, I want this
podcast to be your go-toresource for actionable insights
and inspiring conversations.
Now, if you enjoy the show,please don't forget to subscribe
, leave a review or share itwith your network.
Your support means the worldand really helps bring more
incredible conversations to life.
Justin, thanks for coming on,excited to have you.
Speaker 1 (02:24):
Yeah, great to be here.
Thanks for having me on.
Speaker 2 (02:27):
Absolutely.
I love talking AI, so I'm surethere's a lot that we can get
into today.
Before we jump into things,just for those who are not
familiar, could you give me aquick overview of what Calidus
is and also what the inspirationfor starting the company was?
Speaker 1 (02:42):
Yeah, Calidus is a legal tech platform focused on
work for professionals solawyers, procurement people and
then we're focused on the corelegal and procurement workflows,
not kind of secondary ortertiary work.
Speaker 2 (02:57):
Gotcha.
And then, what was theinspiration for starting it?
Speaker 1 (03:00):
As far as inspiration , I was a lawyer by trade, ended
up doing transformation at AT&Tfor the legal department, and
then, right after that, ran aGen AI group that launched the
first Gen AI product, and itbecame very obvious to me the
work that I was doing intransformation, which was really
impactful, would have beenabsolutely turbocharged if we
(03:21):
had the same Gen AI capabilitiesthat we did a couple of years
later, if we had the same gen-AIcapabilities that we did a
couple of years later, and so itwas just obvious to me that
this was the time to go frommake a transformational impact
on a big industry perspective.
Speaker 2 (03:35):
And congrats, by the way, on making that jump.
That is a big one.
I was curious because, when Iwas checking out your background
, you know, like you mentioned,you'd worked for AT&T.
You'd spent most of your careerprior to starting the
entrepreneurship journey in alarge corporation.
What's the adjustment been likeas you jump from all of that
structure to basically none?
Speaker 1 (03:54):
I love it.
It's one of those where withinevery big corporation, you have
pockets of people that arepretty different from the rest
of the corporation, and so forme it was always being the agent
of change, trying to do thingsfaster, do things in a more
visionary way and make bigtransformations.
And AT&T and DirecTV, where Iwas before great companies,
(04:17):
great people, but it takes solong to move these big
bureaucracies.
My sister she was younger thanme, she had started a startup
and bootstrapped it.
She's done really well withthat.
So I've been pretty helpful orpretty involved with hers for a
bit.
I had done something smallabout 10 years ago, so I've been
aware of the space, payingattention and looking for an
(04:38):
opportunity and eventually onecame up.
So I don't think it was onethat a whole lot of people that
knew me were very surprisedabout that.
My colleagues were definitelylike he's the person trying to
ram, everybody ram stuff through, and he knows that it's going
slow and so it's reallyenjoyable to be able to push
fast.
Speaker 2 (04:59):
Yeah, absolutely.
What did?
What did the ex-colleagues orindustry contacts mentors think
when, when they found out youwere making the jump?
Speaker 1 (05:07):
besides the lack of surprise, um, yeah, definitely,
it was just a.
This is a good fit.
I think it was mostly positive.
Um, as even my boss was like Iknew that we weren't going to
keep you around forever, um,this was, uh, my to go, I think,
and so I still stay in contactwith them, trying to sign them
up as a customer actually rightnow, and it's a great group of
(05:32):
people that I've had a chance towork with at AT&T and their
subsidiary Direct TV.
Speaker 2 (05:36):
So here's a question for you.
So you know, obviously I'vebeen in recruiting for quite
some time now, mostly in thestartup space, and typically the
bias, if you want to call itthat, is that people who come
from big companies are generallynot going to be a fit for
startups because they're used totoo much structure.
There's all the red tape, blah,blah blah, and you mentioned a
(05:56):
little bit.
There's different types ofpeople within these huge
organizations, obviously, butwhat are your thoughts on that
general dynamic?
Would you hire largecorporations for your own
startup?
How are you approaching thatline of thinking?
Speaker 1 (06:11):
I think, generally it's going to be true.
Like I would say most peoplethat I've worked with in the
past, a startup is probably notthe right fit for them.
I think when you look at peopleand see a lot of change in what
they've done, if they've movedaround a lot, that's a good sign
.
It's also a good sign ifthey're in one of the more
transformational groups whetherit's a modernization or
(06:33):
transformation or tech group andthen get a sense of how are
they willing to move fast andare they willing to really align
to the mission and buy into theculture and have a really
different kind of work ethic.
I think those are really thesigns.
It's probably a smallerpercentage of people in bigger
corporations that are willing tomake the move and you don't
want people that they've beentrained to demand like free
(06:58):
coffee all the time and all theperks of working for Google
right, the perks of working forGoogle right.
But you can probably find a fewreally good options if you look
enough and you're open-minded.
Speaker 2 (07:11):
Yeah, yeah, I think that that's a great call out on
the signal of having workedwithin multiple business units
within the business or, yeah, onthe transformation side.
What sort of questions andthat's on paper what sort of
questions would you ask apotential candidate if you were
trying to vet out those morequalitative aspects?
Speaker 1 (07:32):
One.
I try to think about vision andtry to ask them tough
vision-related questions abouthow could we make an impact in
the space.
What I've noticed is,especially in corporate America,
people think pretty smallrelative to in startups.
When I talk to other founders,they're thinking really big, and
if people are willing orthinking about how is the
(07:52):
landscape going to change andhow they're going to capture
value from that and grow theirbusiness in creative ways,
that's a really good sign.
Definitely trying to get asense of can the person wear
multiple hats?
Everybody in a startup has to,even my developers, even though
they're really technical, theyneed to quickly understand the
(08:12):
legal implications for whatthey're building, and so we need
people that arecross-functional, at least in
the way that we do business, andso I'd look for people with
those kind of cross-functionalabilities and just high levels
of curiosity and interest anddrive.
Speaker 2 (08:28):
I love that.
So, to go back to the beginningof the company, you recognize
that there is this need workingin the transformation group.
How did you go from?
There's a need here to actuallycreating and launching the
product, creating and launchingthe product.
Speaker 1 (08:44):
A lot of what we've tried to do is, I mean,
initially, when you zoom out,legal seems like a pretty small
vertical.
It's pretty big though it's atrillion dollars globally in
spend.
There's a lot of ways you canaddress legal meets AI.
We wanted to make something ordo something aggressive,
(09:04):
probably because it's my firsttime I didn't know better that
it had been my fourth time, Iprobably would have gone a lot
smaller.
We went really aggressive andsaid we want to go after the
core workflows.
We know that this is going tobe something where lawyers will
feel uneasy about it at first.
We're going to have to go afterthe early adopters, and so we
try to understand.
(09:24):
We talked to a bunch ofattorneys, we spoke to a lot of
ex-colleagues and all that, andwe got a sense of what are the
things that they spend the mosttime doing, got a really good
understanding of, okay, here'sthe 10 or so big areas, and then
try to map on how could AIreally drive impact in those
different areas?
And then try to think throughokay, what are the problems with
(09:48):
some of these, what are theopportunities with some of these
, and then came up with a listof a couple areas that we were
going to experiment on.
And for us, when we look at usversus some of the other groups,
what we realized pretty quicklywas we had built a group of
developers that were really,really good, and we have
(10:11):
engineers that I would say arebest in the industry right now,
and because of that, we wantedto use that to our advantage.
We've tried to build kind oflike a product superiority
strategy of we're going to buildsuch an amazing product that
it's going to be easier to sellit, versus kind of your
traditional hey, decentpercentage sales and marketing,
decent percentage developmentand so forth.
But by having that approachwe're able to build quickly,
(10:31):
iterate and then learn from thecustomer and learn from actual
usage and then make changes.
And so it's not for everybody,there's other ways to go about
that, but for us that's beenpretty successful and I'm happy
with the approach we're takingon that front.
But for us that's been prettysuccessful and I'm happy with
the approach we're taking onthat front.
Speaker 2 (10:46):
You briefly mentioned earlier that lawyers get a
little bit nervous using AI.
I'm just curious is thatbecause they have their own fear
that they're going to bereplaced?
Is it the fear that the AI isgoing to hallucinate and spit
out something incorrect?
All of the above, somethingelse?
Speaker 1 (10:59):
I think all of the above.
So let's break it down in acouple of different ways.
So one is lawyers do verysophisticated work when you
really look at it.
So think about something thatseems simple contract redlining.
So they're negotiating a dealand they've got maybe a 50-page
document that they've got to gothrough and redline.
There's a lot of things theyneed to take into account as
(11:21):
they think about making changesin the negotiation to this
contract.
So for one, they might have aplaybook where the playbook has
a lot of specific language theywant to reference and they've
got to bring that language inthe playbook.
Two, they've got to understandthe different relationships
between the parties.
What kind of party is it?
What's the market?
What are market terms?
(11:41):
What's the industry they're in?
How much can they push on thesedifferent pieces?
How much does the client careabout this deal?
That's probably a third of thefactors that are really
important of that.
(12:04):
Well, the lawyer sees it andthey're like, okay, great, but I
can't use any of this becauseit's just like a generic red
line.
So this doesn't, I wouldn'tever submit this and my boss
isn't going to accept it and theclient's not going to accept
this.
So there's a lot from theirperspective of you've got to
meet this really high threshold.
At the same time, the hardestpart about running a product is
users, because users have to usethe product and they get
confused really easily.
(12:25):
And the more you build in allthese complex pieces, the easier
it is for them to just kind ofnot understand how it works and
you to create something inelegant.
So you've got to always thinkabout how do we create something
that's elegant and usable whileat the same time, useful for
this end audience.
So I think that's a big part ofwhy lawyers are more reticent
(12:46):
to adopt something like an AItool.
I think on the back end, theyare probably proud people.
They're proud of the work thatthey do and they don't want to
just be told, hey, AI is goingto take your job.
We don't think that's going tohappen anytime remotely soon at
least, Maybe in the far future.
You can talk about that Anytimebefore.
Like artificialsuperintelligence, I think
(13:08):
lawyers are at a point wherethey're going to be people that
AI is going to elevate them andthey're going to just do more
and you're going to have moredemand.
So there's a paradox by aneconomist called Jevin, and so
Jevin's paradox is, as a servicebecomes cheaper, people want
more of it.
You think about supply anddemand, effectively, and it ends
(13:30):
up the result is that thesepeople become more in demand.
As long as they're not as longas somebody else can't just step
in and be a substitute.
In this case, I don't think thatsomeone else, like an end
consumer, is going to take onand do the legal work instead of
the lawyers.
I think it's going to be mostlythe lawyers doing the vast
majority of the work, but the AIis going to help them be a
multiplier and just do more, andthat'll have interesting
(13:53):
implications that some peoplemight be like oh no, we're going
to have so many lawsuits right,and you saw some of that with
Do Not Pay that they're kind ofpushing some pieces there.
That's a reasonable concern, Ithink.
On the other side, though,you'll take the time to make
your contracts very effectiveand write them in a way that's
(14:14):
really thinking about all sortsof liabilities and things that
can happen, and to where youactually start limiting
liability and then, to someextent you don't want I'm going
to.
A lot of us now are in placeswhere, if we get wronged in some
way, like someone justmaterially breaches a contract
or something, you have norecourse because it costs just
(14:36):
too much to sue.
So will that change over time?
I don't know, it might.
Speaker 2 (14:41):
Yeah, that's what I was thinking of, as you were
saying that it's is the the costin order to yeah, exactly, to
sue someone if you've been donewrong in some way.
It's like, well, this personwronged me for three thousand
dollars, but a lawyer is goingto want five, so I guess I'll
just let it go.
So it's it's.
It's really, then, almost thethe barrier to getting that
supply that's going down andthat that, in turn, will get
(15:05):
lawyers more work.
Now I am curious.
It sounds like and correct meif I'm wrong that a lot of the
model or the lawyer populationthat Caledon is servicing is
more on the corporate lawyerside, but then you also have
lawyers that follow the billablemodel.
Right, is there incentive forlawyers to become more efficient
(15:28):
and do more work in less timeif that means a reduction in
their billable hours?
Speaker 1 (15:32):
We do.
We do serve litigators as well,and as far as that question
comes up a lot, I would say thelegal model is going to be slow
to change, but I don't thinkthat that's necessarily terrible
from an AI perspective.
So as an example, when you'reworking for a big law firm, you
might work 3,000 plus hours ayear and you're probably only
(15:54):
billing about 2,200 of thosehours, so about two thirds to
three fourths.
And the reason why is, asyou're doing, kind of simplistic
work or work that just like aclient doesn't want to see on
like a line item you're going towrite it off.
And with AI the hope would bethat you're doing so much, so
efficiently that instead ofworking or billing for 2000
(16:17):
hours, you're billing for almostall of the time that you work.
And then that means for the bigfirms they're going to have
more revenue per attorney andthat's going to be a very
profitable deal for them.
So that's certainly part of thegoal.
The other part would just be youhave attorneys that are
outputting at a really high rateand because of the tools
they're using they're able todemand a higher hourly rate than
(16:40):
attorneys that just refuse touse tools.
And I would bet that theclients and the corporations
that are really driving thelegal bill rates.
They're probably going to seethat and it would become pretty
noticeable because this stuffreally is powerful.
If you're just not using thelatest AI and then one firm is,
they're going to have such a bigadvantage to where everybody's
(17:01):
going to want to use them, towhere I think that it plays out
OK.
And we saw the same thing withwith computers.
Right, it's not like likelawyers are refusing to use a
computer and they're just doingeverything by hand.
The market's eventually goingto push everybody.
It's just going to go a littlebit slower, I think.
Speaker 2 (17:17):
That makes sense, and I know that you mentioned that
you don't anticipate AIreplacing lawyers anytime soon,
unless we get to some sort offuture hypothetical ASI.
What about paralegals, thosewho are doing a little bit more
of the rote work, just the basicfilings that perhaps could be
automated a little bit moreeasily?
Do you see a little bit more ofan automation path there, or
(17:38):
what does that look like?
Speaker 1 (17:43):
an automation path there, or what does that look
like?
It's possible.
My personal opinion is thatthose people oftentimes are
pretty good with technology.
Same with junior attorneys.
They're going to be people whoyou empower with this technology
and then they're really drivingway more efficiency and they're
able to deliver just a lot moreand so well like as we're
selling into the smaller firms.
Usually the first thought isokay, I'm the owner of the place
(18:07):
, I'm kind of the managingpartner.
I usually have to give a lot ofthis work to the associates.
I'm just going to do it myselfnow with the AI, which will
probably give me a better resultthan the associate would.
And the reality is that theassociates love this stuff.
They start using it and they'rejust cranking and they're able
to develop or do things just somuch faster and it's really that
now the partner is just happierwith the results they're
(18:29):
getting.
I would expect that that'sgoing to be more likely in the
near term and it's probably lessof a.
Is it going to be one seniorityperson or another being
replaced?
I think it's more the peoplethat will use AI effectively and
take the time to learn how toare going to be the ones that
demand higher rates and highervalue, and the other groups are
going to just really be lessmarketable very fast.
Speaker 2 (18:56):
That makes sense.
Yeah, I have a buddy.
He's on the recruiting side aswell and he charges hourly and
he charges, I think, probablydouble what most other people do
, and it's basically hisjustification is that it's his
tool set.
You know, he might cost youtwice as much, but he does five
times the work in an hour.
So it's basically the sameprinciple that you're talking
(19:17):
about here.
Speaker 1 (19:18):
Yeah, yeah, absolutely.
And there are some firms thatdo bill on contingency or bill
like a cost per X, like cost perdemand letter, and I think that
those will grow slowly, but butI think it's going to mix up
all the above.
Speaker 2 (19:33):
How do you see AI changing the legal landscape
over?
Call it, the next three to fiveyears.
I think 10 is probably a littletoo far to predict.
Speaker 1 (19:42):
Yeah, yeah.
I think that early on you'llsee we'll hit this kind of S
curve.
So I think the next two yearsor so it's gonna be legal tech
groups like mine are gonna bedeveloping a lot.
They're gonna have, over thenext year or two years, really
hit to the point where trulythey're replacing what some of
(20:03):
the boring work that attorneysdo that is still like
sophisticated but tedious towhere attorneys can focus more
on high value strategic work.
I think that they're going tostart hitting the law firms and
the law firms will see that theyhave to start adopting.
But the adoption cycles aregoing to be slow for the biggest
law firms and when you look atit, just think about it like
(20:25):
you've got these top firms likeSkadden.
They're probably not going tobe have like a specialist who's
the best in the world at somekind of specific form of
litigation in the specificpractice area, be just outright
replaced by AI for what they'redoing today.
That person probably is stilljust going to be an absolute
skilled expert and better thanwhat any near-term AI can do.
(20:48):
At the same time, that personmight have underlying specific
pieces of their work that AI canhelp them do really efficiently
to where they can do the higherlevel thinking, and so, instead
of like trying to replace thatperson's like two week task, it
might be that you're replacingthat person's 20 minute tasks in
(21:09):
doing a lot of that work.
That isn't that.
It's kind of synthesizing a lotof information for them,
boiling it up, and now theydon't have to read every single
case because it's quicklyfinding the information that
they're looking for, orsynthesizing some of the
information in their discoveryand now they can kind of focus
on what really matters.
I think you'll see a lot ofthat in groups that really
understand here's my audience inthe near term are going to be
(21:33):
able to kind of segment theright audience and provide the
right tool for that audience.
The longer term, though, I wouldsay we're probably going to see
some movement in how firmsoperate.
There's some states are alittle bit different now than
they had been in the past.
For example, arizona does notrequire lawyers to own law firms
.
You might see some other statesadopt those principles and
(21:56):
adopt those changes.
If that does happen, you mightsee different ownership models
where PE firms own some of theselaw firms.
They run them a bit differently, but I would expect that you
have the corporations lead theway, pushing for their own
efficiency and then expectingreally high efficiency from
their firms.
(22:16):
And if the firms don't adapt tothese tools, then the corporate
groups are going to bring thework in-house, and so I think
that's adapt to these tools.
Then the corporate groups aregoing to bring the work in-house
, and so I think that's going tokind of push the industry
forward.
And when I kind of zoom out andjust think about what we're
building and our roadmap andwhat really needs to be done, to
where we're really doing a lotof the horror work that's
(22:40):
tedious, it's not going to be along way out, like I think some
people saw GPT 3.5 and they werelike, okay, this is crap, I
can't use this.
We're not there anymore.
We've really come a long wayalready.
The next year or so thesestartups that are really
investing and building quicklyare going to close a lot of the
gap to be really really usefuland I do think that the groups
(23:09):
that adopt are going to be ableto see really big efficiency
gains.
The other piece to layer inthere is there's kind of a
competitive advantage play withautomation.
So if you just automate like10% of the work, you would think
that, okay, I'll be 10% moreeffective.
But it's the same play as whenyou think about comparative
advantage.
You're going to, if 10% of thework is pretty much done,
(23:29):
automatically, the remainingwork that you're doing you're
going to will net out to whereyou get more than a 10%
efficiency gain because you'lljust do, you'll be doing more of
that 10% time.
To where it's building up andtaking on um and just growing
your overall pie.
Speaker 2 (23:48):
Yeah, and then there's also just the reduced
mental task between time,switching focus et cetera.
It's, it's one of those.
One plus one equals threethings.
It sounds like yeah, yeah,that's right, just curious, what
sort of misconceptions are yourunning into frequently from law
, firm owners or variousprospects or customers beyond?
(24:08):
The AI is going to take all ourjobs, thing.
Speaker 1 (24:12):
Yeah, I think part of it is just like under it was
kind of related to it, to that.
It's how.
What is the role of AI?
And for us, the way we break itdown is we try to really
understand what customer we'retrying to serve, because
somebody who's going throughlet's go back to the contract
example a really high volume ofsimple contracts where they just
(24:35):
are looking for the reallycritical issues and just trying
to fix those and maybe they'rein procurement, that's a very
different person to serve, witha very different product, than
somebody that's a verysophisticated lawyer.
And what some groups probablysee is hey, I saw AI that's more
geared toward this procurementgroup and this isn't useful for
(24:56):
me.
Ok, that might be true, butthere is another way that we can
be useful for that skilledattorney, and it's probably more
in the kind of like I mentionedearlier find ways to bubble up
the the kind of smaller tasks towhere they're still in the loop
.
They're still making the endcalls about how things need to
operate and making the strategicdecisions strategic decisions
(25:22):
but they're working incollaboration with an AI that's
doing a lot of the kind ofunderlying routine tasks and
netting to where they can be alot more effective and efficient
.
I think that's just generally amisconception.
I think they think it's like,hey, ai is just going to run
through and just do my contractsfor me and they're like I don't
believe it and it just doesn'tseem like it's capable of it.
I don't see the value.
So that's usually the startingpoint when we first speak with a
(25:45):
law firm or with an attorneyabout AI, and I think once we
show them, no, we're going tokeep you in the loop, we're
going to make everything reallyvisual, really interactive.
Speaker 2 (25:56):
You're still going to be a lot faster and also we're
going to bubble up some thingsyou might've missed and raise
some areas to where we can makethe work quality better, Then
they start becoming prettyinterested.
So as a follow-up question tothat then and you mentioned a
little bit about what theproduct you're building is and
what type of customer it serveswhen you think about your
go-to-market motion, how did youdecide what type of firms to
(26:19):
pursue, what size, et cetera?
Speaker 1 (26:21):
what type of firms to pursue, what size, et cetera.
A lot of it was trial and error.
We had an initial hypothesis.
We changed it a hundred timesand we got a sense of what
customers were using, whatcustomers we were getting the
most traction with and so forth,and what we saw our plan
initially was hey, you've gotreally big ACV with these big
players.
(26:42):
These groups need reallysophisticated AI.
They might want to pay for itand the small players might just
have such low budgets they'rejust not really interested.
I think a lot of groups in theindustry thought that.
And now what we're all seeingis the small law firms are
adopting the fastest and thecorporates probably second, and
(27:04):
the big law firms are reallyslow on the adoption curve and I
expected the change a lot in acouple of years, but for today,
not as much so for us.
We want to meet customers wherethey are.
We don't want to spend a lot oftime just trying to convince a
big firm to buy in a really longsales cycle when we're a
startup and we want to build aproduct that kind of meets the
(27:27):
needs of people and is a reallyuseful product and then scale
that up and go upstream afterwe've done that and so this kind
of fit our focus as well justbe more product focused, really
try to build something useful,get immediate customer feedback,
understand things that arereally the core about how the
product works, rather than kindof tertiary things that a group
(27:51):
might be worried about, and thenkind of go from there as we add
features to scale up.
So we're mostly focused on thesmaller firms and smaller
in-house groups and we've doneit a couple of ways.
So one is we have almost like aD to C motion where we're
looking at go to market, similarto the way that a consumer type
(28:13):
of company would look at it,where you're doing your kind of
SEO and paid search and socialand all that, and then we're
supplementing that with a motionthat's going after small firms
in a sales, more of a salesmotion, and so it's pretty
similar profiles but one sidewill sign up like individual
players.
Then we kind of do like a B2BSaaS motion or bottoms up SaaS
(28:37):
motion and try to kind of landthe rest of the firm.
The other play is more focusedjust from the start.
Let's try to talk to the restof the firm.
The other play is more focusedjust from the start, let's try
to talk to the leaders of thefirm and get them on board.
Speaker 2 (28:47):
Yeah, I was going to ask you about that.
If you're doing PLG or moretraditional outbound motion and
you don't come really from asales background yourself, I'm
just curious, what sort oflearning curve has that been
like?
Are you still doing the salesyourself at this point?
Speaker 1 (29:01):
I still do most of the sales myself.
I'm uh, I work closely withwhat we have one other person
that does sales, but part-timeand and he's great he was a
lawyer for five years at some ofthe top firms.
He left to do sales for alittle bit that he just really
wanted to get into tech and umhe's we've got half of his time
and he does a great job for us.
But then I'll try to raise kindof awareness, do a lot of
(29:24):
content production and then talkto a lot of customers and then
he does kind of some of the rest.
It's been tough, interesting,hard and all that I would say it
was.
One of my biggest learnings is Icame here to build and do
something transformational andtry to kind of establish and use
a vision of what the law lookslike and how it can shift and
(29:46):
change and how a great productcan change that.
And what you realize when youdo a startup is you're selling
24-7 and that's your job.
100% of the time involves asales mindset and so you have to
sell to your team.
You've got to get them excited.
You've got to sell your mission.
You've got to sell to yourinvestors and any future
investors.
(30:07):
You've got to sell to yourcustomers that are existing and
the future ones.
So you're spending all of yourtime doing that and for me it's
been helpful.
I've worked with someone who'slike a coach and like a mentor
and he's been able to kind ofdissect some of the sales
processes that we've had and say, ok, see what you're doing
there, don't do that anymore,let's change and do this, this
(30:30):
and this a little bitdifferently, and that's worked
really well for us.
Speaker 2 (30:34):
Love that.
How are you going to decidewhen it's time to bring on the
first time sales hire, becausethat's obviously such a critical
one?
Speaker 1 (30:44):
We wanted to get to where we're confident.
We've hit product market fitfirst and do founder led sales
up until that point.
Because the feedback loop isreally helpful, that I talk to
customers, I hear theirperspective, I hear their pain
points.
I understand why they think ourproduct's not good enough, and
then we have a really tightfeedback loop with the
(31:06):
developers.
We get stuff out usually thesame day or within the week and
then that really impresses thecustomers and it really just
helps us to build somethingreally useful.
Once we hit that product marketfit perspective, which I think
we're right on the verge of, Ithink that's when you've proven
kind of that sales motion.
I think that's when it makessense to bring somebody else on
(31:28):
full time.
That's in a leadership type ofrole, yep.
Speaker 2 (31:31):
Yep, I think you're generally on the right path
there, and then let's just talkabout the general growth of the
company.
Speaker 1 (31:52):
I know you mentioned that before we got on that you
have about 10 people.
How did you find hire thosepeople?
What was your primarymotivation to raise was how have
you gotten from thatco-founders to the traction
you've gotten so far?
Interesting approach.
I don't know if this is thebest one, the worst one or
somewhere in between, but forour journey it's worth
considering, at least for people.
What we started with was I foundtwo really passionate about AI
guys in the Netherlands and theMiddle East who just really
(32:15):
wanted to build something on aDiscord channel in AI right
after ChatGPT was released andthey were affordable and also
just really wanted to be abuilder, and so I hired them to
build out basically a proof ofconcept and use that proof of
concept plus a bit of vision toraise an angel round from kind
(32:36):
of extended friends and familyof vision to raise an angel
round from kind of extendedfriends and family.
And we had a decent number ofpeople that were interested
because we've done I've beendoing like a real estate
investment group and a lot ofthose group.
The people in the group werelawyers or other people that had
had funds that they wereinterested in investing in AI
and this was a pretty goodvehicle because there wasn't a
whole lot early on of ways youcould invest in kind of the ai
(32:59):
boom, um, just because you caninvest, like in us or google but
google is such like a weirdtype of play right so, um, so
that that helped us and thenthat gave us enough money to
where we could fund realdevelopment and um, and really
grow enough to where we couldget a vc on board and show, show
some customer traction and um,we well, and I brought on our
(33:21):
CTO right after that.
He was somebody who saw like aDiscord job posting and after
I'd been really frustrated bylike everybody that I spoke with
, he saw this like three pageposting that I had that was in a
lot of detail about exactly whowe want.
He responded to every singlelike paragraph or every single
sentence in it just explaining.
(33:42):
Here's exactly why thisperfectly resonates with me and
the two of us just have a verysimilar view on just culture and
speed and the importance ofiteration and just being really
aggressive and all that.
And so he's had just afantastic technical background
where we were just very much inlockstep and it is someone who I
(34:03):
feel like I can perfectlyempower and he can just deliver.
So that worked out pretty wellfor us.
From there, we try to be smartabout the hires, that when
you're a startup, you end upseeing like, ok, we really need
someone to start building thisout quick.
A customer really wants it andwe're doing.
We're really maxed out oncapacity, let's bring somebody
(34:23):
in.
But you have to be deliberate.
I think everybody has heardthis before, but those first few
hires are so important andthey're very likely to set the
culture of the company.
So we really try to bring onpeople that really could cut it
and we're really going to fitthe right kind of fit the
cultural mindset we had andreally understand how to build
(34:44):
in the AI world.
And so we had pretty tough kindof case type questions, had
them think through differentproblems, solve some of them
live or some of them justverbally, and with that I think
we ended up getting some reallygood picks and we've still got
most all the team today foranybody that we've hired.
That's fantastic.
Speaker 2 (35:05):
So you're at 10.
Now.
What do your plans look like,not just from a headcount
perspective, but just generally,for the growth of the business
over the next six months, a yearor two years?
Speaker 1 (35:16):
I don't want to be one of those companies that
grows just to grow people, and Ithink if we were to grow
headcount a ton on the developerside, we'd start seeing some
operational inefficiencies where, like, the team size right now
is pretty efficient and it kindof forces us to think about how
can we solve the problem theeasiest possible way, rather
(35:37):
than like here's, like the big,if we like waterfall it out.
Here's how we we solve theproblem the easiest possible way
, rather than like here's, likethe big, if we like waterfall it
out.
Here's how we might solve it.
And so I don't want to like seea little bit of growth and then
hire three more people on thedeveloper side.
We do need some more salessupport, so eventually we'll
bring someone on there and thenthere's probably gonna be a
(35:58):
little bit of growth.
Beyond that, maybe some helpwith some of the routine tasks
to where I can focus more onproduct and the customer.
But as much as I can, I want toautomate that.
This is a personal opinion, but,being part of a big company,
the kind of trend of talkingabout founder mode resonated
(36:20):
pretty well with me, where, as afounder, let's give like the
most ridiculous example, whereit still makes sense.
So, like on support, I think alot of people will have support
be something that they outsource.
But, like when I look at like acustomer complained about
something and I got an emailabout it in our email, I'll see
(36:42):
it.
I'll immediately know where inthe product the problem is
stemming from.
I talk to the developerdirectly, we talk, we figure out
exactly what the problem is andhow to solve it, and then the
developer releases somethingright away and then I can tell
the customer exactly what we'redoing to fix it and then I can
give them free credits or somekind of refund or something like
(37:04):
that and the problem is solved.
If we had a support person itwould take.
They just realisticallywouldn't care enough to take the
time to really understand theproduct.
They'd probably go through likea longer channel to diagnose
these issues and it just slowsdown the business.
So eventually you have to scaleright, you have to bring in
(37:25):
people, but to the greatestextent possible I'm trying to
run as automated a shop as wecan and use some of these new
tools to do some of theseunderlying functions and then
just take on a little bit moreof the work.
That's going to be morecross-functional.
Speaker 2 (37:40):
Yeah, I think that's a great approach.
Like you said, eventuallythere's going to be a point
where you can't get away with it, but until you reach that point
and you don't have to introducejust extra steps or lack of
feedback loops, then I thinkthat's the right way to do it.
So, final question then, sinceyou mentioned that the founder
mode when you were in the bigcorp resonated.
There's probably going to besome people who are in big
(38:01):
corporations who listen to thisand feel inspired by your own
moves.
So I'm just curious what advicewould you give somebody who
wants to make the leap fromcorporate to startup leadership
or even starting their own thing?
Speaker 1 (38:14):
Yeah, a couple of things.
What worked for me at least wasI jumped around a lot.
I did.
I was a lawyer, was in supplychain, worked for a consulting
firm, worked in an e-com group,did transformation, did a bunch
of finance stuff and then ran anengineering team.
And having that mix ofbackgrounds you don't need to go
(38:36):
that extreme.
That sounds kind of crazy, butjumping around a few times is
really helpful in helping youjust understand all the
different ways a companyoperates and you'll start
picking up, like when I'm inmarketing and I understand the
finance component, I understandthe vendor component.
It just helps me run thatbetter and then it helps you
(38:57):
kind of zoom out big picturewhen you're kind of thinking
about the whole company a lotbetter.
So I think that's one.
The other thing is that thesethe strategy parts of the
corporate when, like thecorporations discussing strategy
at like your big off sites orwhatever, those can be pretty
helpful, trying to just say likereally adopt like a mindset of,
(39:21):
okay, what are we doing?
Well, what are where does thisstrategy not work?
What are the threats?
And being able to not just pareit back the strategy that your
CEO told you, but to reallyrespond to really tough
questions about it and reallythink through the vision.
I think that that's reallyhelpful.
What we did a good job of comingup with at least a high level
strategy at AT&T.
I don't think we executed it aswell as we could have, but our
(39:42):
strategy was always really goodand being able to kind of think
through the elements there wasreally helpful.
And then the last piece is yougot to pay a little bit of
attention to the startup space.
I learned a lot doing this andreally jumping in and I had a
bit of experience before.
If you just go in totally cold,it's pretty hard.
So podcasts like yours are agreat avenue.
(40:05):
Just getting involved in thecommunities, being close with a
startup, maybe as an advisor orsomething similar, would be
really helpful.
But try to just gain that kindof closer, first-hand experience
.
Speaker 2 (40:24):
I love that.
Well, justin, this has beensuper fun.
Real appreciate you coming on.
Thanks for joining Greatspeaking with you.
Bye.
Thank you for tuning in to thisepisode of Seat to Exit.
I hope you found today'sconversation insightful and
valuable.
If you enjoyed the episode,please take a moment to
subscribe, leave a review andshare it with your network.
Your support means the worldhelps us continue to grow and
(40:45):
bring more incredible guestsonto the show.
Now for more content andupdates, follow me on LinkedIn
or Twitter, or you can check outMindHire, where we help
startups build exceptional teams.
Thanks again for listening andI'll see you in the next episode
of Seed to Exit.
When I try to think about vision and try to ask them
tough vision related questionsabout how could we make an
impact on the space, what I'venoticed is, especially in
corporate America, people thinkpretty small relative to in
startups.
When I talk to other founders,they're thinking really big, and
if people are thinking abouthow is the landscape going to
(00:21):
change and how they're going tocapture value from that and grow
their business in creative ways, that's a really good sign.
Definitely trying to get asense of can the person wear
multiple hats?
Everybody in a startup has to,even my developers.
Even though they're reallytechnical, they need to quickly
understand the legalimplications for what they're
(00:42):
building, and so we need peoplethat are cross-functional, um,
at least in the way that we dobusiness, and so I'd look for
people with those kind ofcross-functional abilities and
just high levels of curiosityand interest and drive.
Speaker 2 (00:57):
Thanks for listening.
To Seed to Exit.
I'm Rhys Keck and today I'mjoined by Justin McCallan,
founder and CEO of Calidus AI.
Calidus is an AI companydedicated to augmenting not
replacing, lawyers throughadvanced AI solutions.
Prior to founding Calidus,justin practiced law as a
commercial restructuring and M&Aattorney.
He then spent 12 years at AT&Tand its subsidiaries, including
Zander and DirecTV, includingwhere he launched the first
(01:20):
generative AI product forDirecTV.
In our conversation, we'regoing to talk about the
inspiration for Calidus, what itwas like to transition from a
large enterprise to a startup,the implications of AI in the
legal profession, fundraisinginsights and some advice for
inspiring entrepreneurs.
Really excited for you tolisten.
Let's go ahead and dive intothe episode.
Welcome to Seed to Exit, thepodcast where we uncover the
(01:42):
stories, strategies and insightsthat power the startup
ecosystem.
I'm your host, rhys Keck,founder of MindHire, a talent
acquisition firm specializing inhelping startups build
exceptional teams.
Each week, I sit down withfounders, investors and industry
leaders to explore the journeysbehind iconic companies and
game-changing ideas.
Whether you're building,investing or just curious about
(02:04):
what it takes to succeed in thestartup world, I want this
podcast to be your go-toresource for actionable insights
and inspiring conversations.
Now, if you enjoy the show,please don't forget to subscribe
, leave a review or share itwith your network.
Your support means the worldand really helps bring more
incredible conversations to life.
Justin, thanks for coming on,excited to have you.
Speaker 1 (02:24):
Yeah, great to be here.
Thanks for having me on.
Speaker 2 (02:27):
Absolutely.
I love talking AI, so I'm surethere's a lot that we can get
into today.
Before we jump into things,just for those who are not
familiar, could you give me aquick overview of what Calidus
is and also what the inspirationfor starting the company was?
Speaker 1 (02:42):
Yeah, Calidus is a legal tech platform focused on
work for professionals solawyers, procurement people and
then we're focused on the corelegal and procurement workflows,
not kind of secondary ortertiary work.
Speaker 2 (02:57):
Gotcha.
And then, what was theinspiration for starting it?
Speaker 1 (03:00):
As far as inspiration , I was a lawyer by trade, ended
up doing transformation at AT&Tfor the legal department, and
then, right after that, ran aGen AI group that launched the
first Gen AI product, and itbecame very obvious to me the
work that I was doing intransformation, which was really
impactful, would have beenabsolutely turbocharged if we
(03:21):
had the same Gen AI capabilitiesthat we did a couple of years
later, if we had the same gen-AIcapabilities that we did a
couple of years later, and so itwas just obvious to me that
this was the time to go frommake a transformational impact
on a big industry perspective.
Speaker 2 (03:35):
And congrats, by the way, on making that jump.
That is a big one.
I was curious because, when Iwas checking out your background
, you know, like you mentioned,you'd worked for AT&T.
You'd spent most of your careerprior to starting the
entrepreneurship journey in alarge corporation.
What's the adjustment been likeas you jump from all of that
structure to basically none?
Speaker 1 (03:54):
I love it.
It's one of those where withinevery big corporation, you have
pockets of people that arepretty different from the rest
of the corporation, and so forme it was always being the agent
of change, trying to do thingsfaster, do things in a more
visionary way and make bigtransformations.
And AT&T and DirecTV, where Iwas before great companies,
(04:17):
great people, but it takes solong to move these big
bureaucracies.
My sister she was younger thanme, she had started a startup
and bootstrapped it.
She's done really well withthat.
So I've been pretty helpful orpretty involved with hers for a
bit.
I had done something smallabout 10 years ago, so I've been
aware of the space, payingattention and looking for an
(04:38):
opportunity and eventually onecame up.
So I don't think it was onethat a whole lot of people that
knew me were very surprisedabout that.
My colleagues were definitelylike he's the person trying to
ram, everybody ram stuff through, and he knows that it's going
slow and so it's reallyenjoyable to be able to push
fast.
Speaker 2 (04:59):
Yeah, absolutely.
What did?
What did the ex-colleagues orindustry contacts mentors think
when, when they found out youwere making the jump?
Speaker 1 (05:07):
besides the lack of surprise, um, yeah, definitely,
it was just a.
This is a good fit.
I think it was mostly positive.
Um, as even my boss was like Iknew that we weren't going to
keep you around forever, um,this was, uh, my to go, I think,
and so I still stay in contactwith them, trying to sign them
up as a customer actually rightnow, and it's a great group of
(05:32):
people that I've had a chance towork with at AT&T and their
subsidiary Direct TV.
Speaker 2 (05:36):
So here's a question for you.
So you know, obviously I'vebeen in recruiting for quite
some time now, mostly in thestartup space, and typically the
bias, if you want to call itthat, is that people who come
from big companies are generallynot going to be a fit for
startups because they're used totoo much structure.
There's all the red tape, blah,blah blah, and you mentioned a
(05:56):
little bit.
There's different types ofpeople within these huge
organizations, obviously, butwhat are your thoughts on that
general dynamic?
Would you hire largecorporations for your own
startup?
How are you approaching thatline of thinking?
Speaker 1 (06:11):
I think, generally it's going to be true.
Like I would say most peoplethat I've worked with in the
past, a startup is probably notthe right fit for them.
I think when you look at peopleand see a lot of change in what
they've done, if they've movedaround a lot, that's a good sign
.
It's also a good sign ifthey're in one of the more
transformational groups whetherit's a modernization or
(06:33):
transformation or tech group andthen get a sense of how are
they willing to move fast andare they willing to really align
to the mission and buy into theculture and have a really
different kind of work ethic.
I think those are really thesigns.
It's probably a smallerpercentage of people in bigger
corporations that are willing tomake the move and you don't
want people that they've beentrained to demand like free
(06:58):
coffee all the time and all theperks of working for Google
right, the perks of working forGoogle right.
But you can probably find a fewreally good options if you look
enough and you're open-minded.
Speaker 2 (07:11):
Yeah, yeah, I think that that's a great call out on
the signal of having workedwithin multiple business units
within the business or, yeah, onthe transformation side.
What sort of questions andthat's on paper what sort of
questions would you ask apotential candidate if you were
trying to vet out those morequalitative aspects?
Speaker 1 (07:32):
One.
I try to think about vision andtry to ask them tough
vision-related questions abouthow could we make an impact in
the space.
What I've noticed is,especially in corporate America,
people think pretty smallrelative to in startups.
When I talk to other founders,they're thinking really big, and
if people are willing orthinking about how is the
(07:52):
landscape going to change andhow they're going to capture
value from that and grow theirbusiness in creative ways,
that's a really good sign.
Definitely trying to get asense of can the person wear
multiple hats?
Everybody in a startup has to,even my developers, even though
they're really technical, theyneed to quickly understand the
(08:12):
legal implications for whatthey're building, and so we need
people that arecross-functional, at least in
the way that we do business, andso I'd look for people with
those kind of cross-functionalabilities and just high levels
of curiosity and interest anddrive.
Speaker 2 (08:28):
I love that.
So, to go back to the beginningof the company, you recognize
that there is this need workingin the transformation group.
How did you go from?
There's a need here to actuallycreating and launching the
product, creating and launchingthe product.
Speaker 1 (08:44):
A lot of what we've tried to do is, I mean,
initially, when you zoom out,legal seems like a pretty small
vertical.
It's pretty big though it's atrillion dollars globally in
spend.
There's a lot of ways you canaddress legal meets AI.
We wanted to make something ordo something aggressive,
(09:04):
probably because it's my firsttime I didn't know better that
it had been my fourth time, Iprobably would have gone a lot
smaller.
We went really aggressive andsaid we want to go after the
core workflows.
We know that this is going tobe something where lawyers will
feel uneasy about it at first.
We're going to have to go afterthe early adopters, and so we
try to understand.
(09:24):
We talked to a bunch ofattorneys, we spoke to a lot of
ex-colleagues and all that, andwe got a sense of what are the
things that they spend the mosttime doing, got a really good
understanding of, okay, here'sthe 10 or so big areas, and then
try to map on how could AIreally drive impact in those
different areas?
And then try to think throughokay, what are the problems with
(09:48):
some of these, what are theopportunities with some of these
, and then came up with a listof a couple areas that we were
going to experiment on.
And for us, when we look at usversus some of the other groups,
what we realized pretty quicklywas we had built a group of
developers that were really,really good, and we have
(10:11):
engineers that I would say arebest in the industry right now,
and because of that, we wantedto use that to our advantage.
We've tried to build kind oflike a product superiority
strategy of we're going to buildsuch an amazing product that
it's going to be easier to sellit, versus kind of your
traditional hey, decentpercentage sales and marketing,
decent percentage developmentand so forth.
But by having that approachwe're able to build quickly,
(10:31):
iterate and then learn from thecustomer and learn from actual
usage and then make changes.
And so it's not for everybody,there's other ways to go about
that, but for us that's beenpretty successful and I'm happy
with the approach we're takingon that front.
But for us that's been prettysuccessful and I'm happy with
the approach we're taking onthat front.
Speaker 2 (10:46):
You briefly mentioned earlier that lawyers get a
little bit nervous using AI.
I'm just curious is thatbecause they have their own fear
that they're going to bereplaced?
Is it the fear that the AI isgoing to hallucinate and spit
out something incorrect?
All of the above, somethingelse?
Speaker 1 (10:59):
I think all of the above.
So let's break it down in acouple of different ways.
So one is lawyers do verysophisticated work when you
really look at it.
So think about something thatseems simple contract redlining.
So they're negotiating a dealand they've got maybe a 50-page
document that they've got to gothrough and redline.
There's a lot of things theyneed to take into account as
(11:21):
they think about making changesin the negotiation to this
contract.
So for one, they might have aplaybook where the playbook has
a lot of specific language theywant to reference and they've
got to bring that language inthe playbook.
Two, they've got to understandthe different relationships
between the parties.
What kind of party is it?
What's the market?
What are market terms?
(11:41):
What's the industry they're in?
How much can they push on thesedifferent pieces?
How much does the client careabout this deal?
That's probably a third of thefactors that are really
important of that.
(12:04):
Well, the lawyer sees it andthey're like, okay, great, but I
can't use any of this becauseit's just like a generic red
line.
So this doesn't, I wouldn'tever submit this and my boss
isn't going to accept it and theclient's not going to accept
this.
So there's a lot from theirperspective of you've got to
meet this really high threshold.
At the same time, the hardestpart about running a product is
users, because users have to usethe product and they get
confused really easily.
(12:25):
And the more you build in allthese complex pieces, the easier
it is for them to just kind ofnot understand how it works and
you to create something inelegant.
So you've got to always thinkabout how do we create something
that's elegant and usable whileat the same time, useful for
this end audience.
So I think that's a big part ofwhy lawyers are more reticent
(12:46):
to adopt something like an AItool.
I think on the back end, theyare probably proud people.
They're proud of the work thatthey do and they don't want to
just be told, hey, AI is goingto take your job.
We don't think that's going tohappen anytime remotely soon at
least, Maybe in the far future.
You can talk about that Anytimebefore.
Like artificialsuperintelligence, I think
(13:08):
lawyers are at a point wherethey're going to be people that
AI is going to elevate them andthey're going to just do more
and you're going to have moredemand.
So there's a paradox by aneconomist called Jevin, and so
Jevin's paradox is, as a servicebecomes cheaper, people want
more of it.
You think about supply anddemand, effectively, and it ends
(13:30):
up the result is that thesepeople become more in demand.
As long as they're not as longas somebody else can't just step
in and be a substitute.
In this case, I don't think thatsomeone else, like an end
consumer, is going to take onand do the legal work instead of
the lawyers.
I think it's going to be mostlythe lawyers doing the vast
majority of the work, but the AIis going to help them be a
multiplier and just do more, andthat'll have interesting
(13:53):
implications that some peoplemight be like oh no, we're going
to have so many lawsuits right,and you saw some of that with
Do Not Pay that they're kind ofpushing some pieces there.
That's a reasonable concern, Ithink.
On the other side, though,you'll take the time to make
your contracts very effectiveand write them in a way that's
(14:14):
really thinking about all sortsof liabilities and things that
can happen, and to where youactually start limiting
liability and then, to someextent you don't want I'm going
to.
A lot of us now are in placeswhere, if we get wronged in some
way, like someone justmaterially breaches a contract
or something, you have norecourse because it costs just
(14:36):
too much to sue.
So will that change over time?
I don't know, it might.
Speaker 2 (14:41):
Yeah, that's what I was thinking of, as you were
saying that it's is the the costin order to yeah, exactly, to
sue someone if you've been donewrong in some way.
It's like, well, this personwronged me for three thousand
dollars, but a lawyer is goingto want five, so I guess I'll
just let it go.
So it's it's.
It's really, then, almost thethe barrier to getting that
supply that's going down andthat that, in turn, will get
(15:05):
lawyers more work.
Now I am curious.
It sounds like and correct meif I'm wrong that a lot of the
model or the lawyer populationthat Caledon is servicing is
more on the corporate lawyerside, but then you also have
lawyers that follow the billablemodel.
Right, is there incentive forlawyers to become more efficient
(15:28):
and do more work in less timeif that means a reduction in
their billable hours?
Speaker 1 (15:32):
We do.
We do serve litigators as well,and as far as that question
comes up a lot, I would say thelegal model is going to be slow
to change, but I don't thinkthat that's necessarily terrible
from an AI perspective.
So as an example, when you'reworking for a big law firm, you
might work 3,000 plus hours ayear and you're probably only
(15:54):
billing about 2,200 of thosehours, so about two thirds to
three fourths.
And the reason why is, asyou're doing, kind of simplistic
work or work that just like aclient doesn't want to see on
like a line item you're going towrite it off.
And with AI the hope would bethat you're doing so much, so
efficiently that instead ofworking or billing for 2000
(16:17):
hours, you're billing for almostall of the time that you work.
And then that means for the bigfirms they're going to have
more revenue per attorney andthat's going to be a very
profitable deal for them.
So that's certainly part of thegoal.
The other part would just be youhave attorneys that are
outputting at a really high rateand because of the tools
they're using they're able todemand a higher hourly rate than
(16:40):
attorneys that just refuse touse tools.
And I would bet that theclients and the corporations
that are really driving thelegal bill rates.
They're probably going to seethat and it would become pretty
noticeable because this stuffreally is powerful.
If you're just not using thelatest AI and then one firm is,
they're going to have such a bigadvantage to where everybody's
(17:01):
going to want to use them, towhere I think that it plays out
OK.
And we saw the same thing withwith computers.
Right, it's not like likelawyers are refusing to use a
computer and they're just doingeverything by hand.
The market's eventually goingto push everybody.
It's just going to go a littlebit slower, I think.
Speaker 2 (17:17):
That makes sense, and I know that you mentioned that
you don't anticipate AIreplacing lawyers anytime soon,
unless we get to some sort offuture hypothetical ASI.
What about paralegals, thosewho are doing a little bit more
of the rote work, just the basicfilings that perhaps could be
automated a little bit moreeasily?
Do you see a little bit more ofan automation path there, or
(17:38):
what does that look like?
Speaker 1 (17:43):
an automation path there, or what does that look
like?
It's possible.
My personal opinion is thatthose people oftentimes are
pretty good with technology.
Same with junior attorneys.
They're going to be people whoyou empower with this technology
and then they're really drivingway more efficiency and they're
able to deliver just a lot moreand so well like as we're
selling into the smaller firms.
Usually the first thought isokay, I'm the owner of the place
(18:07):
, I'm kind of the managingpartner.
I usually have to give a lot ofthis work to the associates.
I'm just going to do it myselfnow with the AI, which will
probably give me a better resultthan the associate would.
And the reality is that theassociates love this stuff.
They start using it and they'rejust cranking and they're able
to develop or do things just somuch faster and it's really that
now the partner is just happierwith the results they're
(18:29):
getting.
I would expect that that'sgoing to be more likely in the
near term and it's probably lessof a.
Is it going to be one seniorityperson or another being
replaced?
I think it's more the peoplethat will use AI effectively and
take the time to learn how toare going to be the ones that
demand higher rates and highervalue, and the other groups are
going to just really be lessmarketable very fast.
Speaker 2 (18:56):
That makes sense.
Yeah, I have a buddy.
He's on the recruiting side aswell and he charges hourly and
he charges, I think, probablydouble what most other people do
, and it's basically hisjustification is that it's his
tool set.
You know, he might cost youtwice as much, but he does five
times the work in an hour.
So it's basically the sameprinciple that you're talking
(19:17):
about here.
Speaker 1 (19:18):
Yeah, yeah, absolutely.
And there are some firms thatdo bill on contingency or bill
like a cost per X, like cost perdemand letter, and I think that
those will grow slowly, but butI think it's going to mix up
all the above.
Speaker 2 (19:33):
How do you see AI changing the legal landscape
over?
Call it, the next three to fiveyears.
I think 10 is probably a littletoo far to predict.
Speaker 1 (19:42):
Yeah, yeah.
I think that early on you'llsee we'll hit this kind of S
curve.
So I think the next two yearsor so it's gonna be legal tech
groups like mine are gonna bedeveloping a lot.
They're gonna have, over thenext year or two years, really
hit to the point where trulythey're replacing what some of
(20:03):
the boring work that attorneysdo that is still like
sophisticated but tedious towhere attorneys can focus more
on high value strategic work.
I think that they're going tostart hitting the law firms and
the law firms will see that theyhave to start adopting.
But the adoption cycles aregoing to be slow for the biggest
law firms and when you look atit, just think about it like
(20:25):
you've got these top firms likeSkadden.
They're probably not going tobe have like a specialist who's
the best in the world at somekind of specific form of
litigation in the specificpractice area, be just outright
replaced by AI for what they'redoing today.
That person probably is stilljust going to be an absolute
skilled expert and better thanwhat any near-term AI can do.
(20:48):
At the same time, that personmight have underlying specific
pieces of their work that AI canhelp them do really efficiently
to where they can do the higherlevel thinking, and so, instead
of like trying to replace thatperson's like two week task, it
might be that you're replacingthat person's 20 minute tasks in
(21:09):
doing a lot of that work.
That isn't that.
It's kind of synthesizing a lotof information for them,
boiling it up, and now theydon't have to read every single
case because it's quicklyfinding the information that
they're looking for, orsynthesizing some of the
information in their discoveryand now they can kind of focus
on what really matters.
I think you'll see a lot ofthat in groups that really
understand here's my audience inthe near term are going to be
(21:33):
able to kind of segment theright audience and provide the
right tool for that audience.
The longer term, though, I wouldsay we're probably going to see
some movement in how firmsoperate.
There's some states are alittle bit different now than
they had been in the past.
For example, arizona does notrequire lawyers to own law firms
.
You might see some other statesadopt those principles and
(21:56):
adopt those changes.
If that does happen, you mightsee different ownership models
where PE firms own some of theselaw firms.
They run them a bit differently, but I would expect that you
have the corporations lead theway, pushing for their own
efficiency and then expectingreally high efficiency from
their firms.
(22:16):
And if the firms don't adapt tothese tools, then the corporate
groups are going to bring thework in-house, and so I think
that's adapt to these tools.
Then the corporate groups aregoing to bring the work in-house
, and so I think that's going tokind of push the industry
forward.
And when I kind of zoom out andjust think about what we're
building and our roadmap andwhat really needs to be done, to
where we're really doing a lotof the horror work that's
(22:40):
tedious, it's not going to be along way out, like I think some
people saw GPT 3.5 and they werelike, okay, this is crap, I
can't use this.
We're not there anymore.
We've really come a long wayalready.
The next year or so thesestartups that are really
investing and building quicklyare going to close a lot of the
gap to be really really usefuland I do think that the groups
(23:09):
that adopt are going to be ableto see really big efficiency
gains.
The other piece to layer inthere is there's kind of a
competitive advantage play withautomation.
So if you just automate like10% of the work, you would think
that, okay, I'll be 10% moreeffective.
But it's the same play as whenyou think about comparative
advantage.
You're going to, if 10% of thework is pretty much done,
(23:29):
automatically, the remainingwork that you're doing you're
going to will net out to whereyou get more than a 10%
efficiency gain because you'lljust do, you'll be doing more of
that 10% time.
To where it's building up andtaking on um and just growing
your overall pie.
Speaker 2 (23:48):
Yeah, and then there's also just the reduced
mental task between time,switching focus et cetera.
It's, it's one of those.
One plus one equals threethings.
It sounds like yeah, yeah,that's right, just curious, what
sort of misconceptions are yourunning into frequently from law
, firm owners or variousprospects or customers beyond?
(24:08):
The AI is going to take all ourjobs, thing.
Speaker 1 (24:12):
Yeah, I think part of it is just like under it was
kind of related to it, to that.
It's how.
What is the role of AI?
And for us, the way we break itdown is we try to really
understand what customer we'retrying to serve, because
somebody who's going throughlet's go back to the contract
example a really high volume ofsimple contracts where they just
(24:35):
are looking for the reallycritical issues and just trying
to fix those and maybe they'rein procurement, that's a very
different person to serve, witha very different product, than
somebody that's a verysophisticated lawyer.
And what some groups probablysee is hey, I saw AI that's more
geared toward this procurementgroup and this isn't useful for
(24:56):
me.
Ok, that might be true, butthere is another way that we can
be useful for that skilledattorney, and it's probably more
in the kind of like I mentionedearlier find ways to bubble up
the the kind of smaller tasks towhere they're still in the loop
.
They're still making the endcalls about how things need to
operate and making the strategicdecisions strategic decisions
(25:22):
but they're working incollaboration with an AI that's
doing a lot of the kind ofunderlying routine tasks and
netting to where they can be alot more effective and efficient
.
I think that's just generally amisconception.
I think they think it's like,hey, ai is just going to run
through and just do my contractsfor me and they're like I don't
believe it and it just doesn'tseem like it's capable of it.
I don't see the value.
So that's usually the startingpoint when we first speak with a
(25:45):
law firm or with an attorneyabout AI, and I think once we
show them, no, we're going tokeep you in the loop, we're
going to make everything reallyvisual, really interactive.
Speaker 2 (25:56):
You're still going to be a lot faster and also we're
going to bubble up some thingsyou might've missed and raise
some areas to where we can makethe work quality better, Then
they start becoming prettyinterested.
So as a follow-up question tothat then and you mentioned a
little bit about what theproduct you're building is and
what type of customer it serveswhen you think about your
go-to-market motion, how did youdecide what type of firms to
(26:19):
pursue, what size, et cetera?
Speaker 1 (26:21):
what type of firms to pursue, what size, et cetera.
A lot of it was trial and error.
We had an initial hypothesis.
We changed it a hundred timesand we got a sense of what
customers were using, whatcustomers we were getting the
most traction with and so forth,and what we saw our plan
initially was hey, you've gotreally big ACV with these big
players.
(26:42):
These groups need reallysophisticated AI.
They might want to pay for itand the small players might just
have such low budgets they'rejust not really interested.
I think a lot of groups in theindustry thought that.
And now what we're all seeingis the small law firms are
adopting the fastest and thecorporates probably second, and
(27:04):
the big law firms are reallyslow on the adoption curve and I
expected the change a lot in acouple of years, but for today,
not as much so for us.
We want to meet customers wherethey are.
We don't want to spend a lot oftime just trying to convince a
big firm to buy in a really longsales cycle when we're a
startup and we want to build aproduct that kind of meets the
(27:27):
needs of people and is a reallyuseful product and then scale
that up and go upstream afterwe've done that and so this kind
of fit our focus as well justbe more product focused, really
try to build something useful,get immediate customer feedback,
understand things that arereally the core about how the
product works, rather than kindof tertiary things that a group
(27:51):
might be worried about, and thenkind of go from there as we add
features to scale up.
So we're mostly focused on thesmaller firms and smaller
in-house groups and we've doneit a couple of ways.
So one is we have almost like aD to C motion where we're
looking at go to market, similarto the way that a consumer type
(28:13):
of company would look at it,where you're doing your kind of
SEO and paid search and socialand all that, and then we're
supplementing that with a motionthat's going after small firms
in a sales, more of a salesmotion, and so it's pretty
similar profiles but one sidewill sign up like individual
players.
Then we kind of do like a B2BSaaS motion or bottoms up SaaS
(28:37):
motion and try to kind of landthe rest of the firm.
The other play is more focusedjust from the start.
Let's try to talk to the restof the firm.
The other play is more focusedjust from the start, let's try
to talk to the leaders of thefirm and get them on board.
Speaker 2 (28:47):
Yeah, I was going to ask you about that.
If you're doing PLG or moretraditional outbound motion and
you don't come really from asales background yourself, I'm
just curious, what sort oflearning curve has that been
like?
Are you still doing the salesyourself at this point?
Speaker 1 (29:01):
I still do most of the sales myself.
I'm uh, I work closely withwhat we have one other person
that does sales, but part-timeand and he's great he was a
lawyer for five years at some ofthe top firms.
He left to do sales for alittle bit that he just really
wanted to get into tech and umhe's we've got half of his time
and he does a great job for us.
But then I'll try to raise kindof awareness, do a lot of
(29:24):
content production and then talkto a lot of customers and then
he does kind of some of the rest.
It's been tough, interesting,hard and all that I would say it
was.
One of my biggest learnings is Icame here to build and do
something transformational andtry to kind of establish and use
a vision of what the law lookslike and how it can shift and
(29:46):
change and how a great productcan change that.
And what you realize when youdo a startup is you're selling
24-7 and that's your job.
100% of the time involves asales mindset and so you have to
sell to your team.
You've got to get them excited.
You've got to sell your mission.
You've got to sell to yourinvestors and any future
investors.
(30:07):
You've got to sell to yourcustomers that are existing and
the future ones.
So you're spending all of yourtime doing that and for me it's
been helpful.
I've worked with someone who'slike a coach and like a mentor
and he's been able to kind ofdissect some of the sales
processes that we've had and say, ok, see what you're doing
there, don't do that anymore,let's change and do this, this
(30:30):
and this a little bitdifferently, and that's worked
really well for us.
Speaker 2 (30:34):
Love that.
How are you going to decidewhen it's time to bring on the
first time sales hire, becausethat's obviously such a critical
one?
Speaker 1 (30:44):
We wanted to get to where we're confident.
We've hit product market fitfirst and do founder led sales
up until that point.
Because the feedback loop isreally helpful, that I talk to
customers, I hear theirperspective, I hear their pain
points.
I understand why they think ourproduct's not good enough, and
then we have a really tightfeedback loop with the
(31:06):
developers.
We get stuff out usually thesame day or within the week and
then that really impresses thecustomers and it really just
helps us to build somethingreally useful.
Once we hit that product marketfit perspective, which I think
we're right on the verge of, Ithink that's when you've proven
kind of that sales motion.
I think that's when it makessense to bring somebody else on
(31:28):
full time.
That's in a leadership type ofrole, yep.
Speaker 2 (31:31):
Yep, I think you're generally on the right path
there, and then let's just talkabout the general growth of the
company.
Speaker 1 (31:52):
I know you mentioned that before we got on that you
have about 10 people.
How did you find hire thosepeople?
What was your primarymotivation to raise was how have
you gotten from thatco-founders to the traction
you've gotten so far?
Interesting approach.
I don't know if this is thebest one, the worst one or
somewhere in between, but forour journey it's worth
considering, at least for people.
What we started with was I foundtwo really passionate about AI
guys in the Netherlands and theMiddle East who just really
(32:15):
wanted to build something on aDiscord channel in AI right
after ChatGPT was released andthey were affordable and also
just really wanted to be abuilder, and so I hired them to
build out basically a proof ofconcept and use that proof of
concept plus a bit of vision toraise an angel round from kind
(32:36):
of extended friends and familyof vision to raise an angel
round from kind of extendedfriends and family.
And we had a decent number ofpeople that were interested
because we've done I've beendoing like a real estate
investment group and a lot ofthose group.
The people in the group werelawyers or other people that had
had funds that they wereinterested in investing in AI
and this was a pretty goodvehicle because there wasn't a
whole lot early on of ways youcould invest in kind of the ai
(32:59):
boom, um, just because you caninvest, like in us or google but
google is such like a weirdtype of play right so, um, so
that that helped us and thenthat gave us enough money to
where we could fund realdevelopment and um, and really
grow enough to where we couldget a vc on board and show, show
some customer traction and um,we well, and I brought on our
(33:21):
CTO right after that.
He was somebody who saw like aDiscord job posting and after
I'd been really frustrated bylike everybody that I spoke with
, he saw this like three pageposting that I had that was in a
lot of detail about exactly whowe want.
He responded to every singlelike paragraph or every single
sentence in it just explaining.
(33:42):
Here's exactly why thisperfectly resonates with me and
the two of us just have a verysimilar view on just culture and
speed and the importance ofiteration and just being really
aggressive and all that.
And so he's had just afantastic technical background
where we were just very much inlockstep and it is someone who I
(34:03):
feel like I can perfectlyempower and he can just deliver.
So that worked out pretty wellfor us.
From there, we try to be smartabout the hires, that when
you're a startup, you end upseeing like, ok, we really need
someone to start building thisout quick.
A customer really wants it andwe're doing.
We're really maxed out oncapacity, let's bring somebody
(34:23):
in.
But you have to be deliberate.
I think everybody has heardthis before, but those first few
hires are so important andthey're very likely to set the
culture of the company.
So we really try to bring onpeople that really could cut it
and we're really going to fitthe right kind of fit the
cultural mindset we had andreally understand how to build
(34:44):
in the AI world.
And so we had pretty tough kindof case type questions, had
them think through differentproblems, solve some of them
live or some of them justverbally, and with that I think
we ended up getting some reallygood picks and we've still got
most all the team today foranybody that we've hired.
That's fantastic.
Speaker 2 (35:05):
So you're at 10.
Now.
What do your plans look like,not just from a headcount
perspective, but just generally,for the growth of the business
over the next six months, a yearor two years?
Speaker 1 (35:16):
I don't want to be one of those companies that
grows just to grow people, and Ithink if we were to grow
headcount a ton on the developerside, we'd start seeing some
operational inefficiencies where, like, the team size right now
is pretty efficient and it kindof forces us to think about how
can we solve the problem theeasiest possible way, rather
(35:37):
than like here's, like the big,if we like waterfall it out.
Here's how we we solve theproblem the easiest possible way
, rather than like here's, likethe big, if we like waterfall it
out.
Here's how we might solve it.
And so I don't want to like seea little bit of growth and then
hire three more people on thedeveloper side.
We do need some more salessupport, so eventually we'll
bring someone on there and thenthere's probably gonna be a
(35:58):
little bit of growth.
Beyond that, maybe some helpwith some of the routine tasks
to where I can focus more onproduct and the customer.
But as much as I can, I want toautomate that.
This is a personal opinion, but,being part of a big company,
the kind of trend of talkingabout founder mode resonated
(36:20):
pretty well with me, where, as afounder, let's give like the
most ridiculous example, whereit still makes sense.
So, like on support, I think alot of people will have support
be something that they outsource.
But, like when I look at like acustomer complained about
something and I got an emailabout it in our email, I'll see
(36:42):
it.
I'll immediately know where inthe product the problem is
stemming from.
I talk to the developerdirectly, we talk, we figure out
exactly what the problem is andhow to solve it, and then the
developer releases somethingright away and then I can tell
the customer exactly what we'redoing to fix it and then I can
give them free credits or somekind of refund or something like
(37:04):
that and the problem is solved.
If we had a support person itwould take.
They just realisticallywouldn't care enough to take the
time to really understand theproduct.
They'd probably go through likea longer channel to diagnose
these issues and it just slowsdown the business.
So eventually you have to scaleright, you have to bring in
(37:25):
people, but to the greatestextent possible I'm trying to
run as automated a shop as wecan and use some of these new
tools to do some of theseunderlying functions and then
just take on a little bit moreof the work.
That's going to be morecross-functional.
Speaker 2 (37:40):
Yeah, I think that's a great approach.
Like you said, eventuallythere's going to be a point
where you can't get away with it, but until you reach that point
and you don't have to introducejust extra steps or lack of
feedback loops, then I thinkthat's the right way to do it.
So, final question then, sinceyou mentioned that the founder
mode when you were in the bigcorp resonated.
There's probably going to besome people who are in big
(38:01):
corporations who listen to thisand feel inspired by your own
moves.
So I'm just curious what advicewould you give somebody who
wants to make the leap fromcorporate to startup leadership
or even starting their own thing?
Speaker 1 (38:14):
Yeah, a couple of things.
What worked for me at least wasI jumped around a lot.
I did.
I was a lawyer, was in supplychain, worked for a consulting
firm, worked in an e-com group,did transformation, did a bunch
of finance stuff and then ran anengineering team.
And having that mix ofbackgrounds you don't need to go
(38:36):
that extreme.
That sounds kind of crazy, butjumping around a few times is
really helpful in helping youjust understand all the
different ways a companyoperates and you'll start
picking up, like when I'm inmarketing and I understand the
finance component, I understandthe vendor component.
It just helps me run thatbetter and then it helps you
(38:57):
kind of zoom out big picturewhen you're kind of thinking
about the whole company a lotbetter.
So I think that's one.
The other thing is that thesethe strategy parts of the
corporate when, like thecorporations discussing strategy
at like your big off sites orwhatever, those can be pretty
helpful, trying to just say likereally adopt like a mindset of,
(39:21):
okay, what are we doing?
Well, what are where does thisstrategy not work?
What are the threats?
And being able to not just pareit back the strategy that your
CEO told you, but to reallyrespond to really tough
questions about it and reallythink through the vision.
I think that that's reallyhelpful.
What we did a good job of comingup with at least a high level
strategy at AT&T.
I don't think we executed it aswell as we could have, but our
(39:42):
strategy was always really goodand being able to kind of think
through the elements there wasreally helpful.
And then the last piece is yougot to pay a little bit of
attention to the startup space.
I learned a lot doing this andreally jumping in and I had a
bit of experience before.
If you just go in totally cold,it's pretty hard.
So podcasts like yours are agreat avenue.
(40:05):
Just getting involved in thecommunities, being close with a
startup, maybe as an advisor orsomething similar, would be
really helpful.
But try to just gain that kindof closer, first-hand experience
.
Speaker 2 (40:24):
I love that.
Well, justin, this has beensuper fun.
Real appreciate you coming on.
Thanks for joining Greatspeaking with you.
Bye.
Thank you for tuning in to thisepisode of Seat to Exit.
I hope you found today'sconversation insightful and
valuable.
If you enjoyed the episode,please take a moment to
subscribe, leave a review andshare it with your network.
Your support means the worldhelps us continue to grow and
(40:45):
bring more incredible guestsonto the show.
Now for more content andupdates, follow me on LinkedIn
or Twitter, or you can check outMindHire, where we help
startups build exceptional teams.
Thanks again for listening andI'll see you in the next episode
of Seed to Exit.
Popular Podcasts
True Crime Tonight
If you eat, sleep, and breathe true crime, TRUE CRIME TONIGHT is serving up your nightly fix. Five nights a week, KT STUDIOS & iHEART RADIO invite listeners to pull up a seat for an unfiltered look at the biggest cases making headlines, celebrity scandals, and the trials everyone is watching. With a mix of expert analysis, hot takes, and listener call-ins, TRUE CRIME TONIGHT goes beyond the headlines to uncover the twists, turns, and unanswered questions that keep us all obsessed—because, at TRUE CRIME TONIGHT, there’s a seat for everyone. Whether breaking down crime scene forensics, scrutinizing serial killers, or debating the most binge-worthy true crime docs, True Crime Tonight is the fresh, fast-paced, and slightly addictive home for true crime lovers.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.