Vibecoding Legal Work

I spoke with Matt Pollins, the Co-Founder and Chief Product Officer at Lupl, a work management platform for attorneys and agent teams. He is also the Co-Founder of Vibecode.law, an open community platform launched in 2026 for legal professionals to learn how to build apps and share what they’ve made. We discussed the value of combining attorneys with Agent Teams, the genesis of vibecode.law, and the value of vibecoding in legal.

Episode Transcript

(00:01):
Welcome to Reinventing Professionals,a podcast hosted by industry analyst
Ari Kaplan, which shares ideas,guidance, and perspectives from market
leaders shaping the next generationof legal and professional services.
This is Ari Kaplan and I'm speakingtoday with Matt Pollins, the

(00:23):
co-founder and chief product Officerat Loophole, a work management
platform for attorneys and agent teams.
He is also the co-founder ofVibe Code Law, an open community
platform launched in 2026 for legalprofessionals to learn how to build
apps and share what they've made.
Hi Matt.
How are you?

(00:43):
Hey, Ari, it's good to be here.
Thanks for having me.
Oh, it's my privilege.
I'm looking forward to the conversation.
Tell us about your backgroundand the genesis of Loople.
I am an attorney by background.
I spent about half my career in Londonand the other half out in Singapore.
I went from being a tech lawyer.
Working with companies like Microsoftand Meta and the big tech companies

(01:05):
as well as startups to legal tech.
So pivoted really from practicingto co-founding loophole.
Around 2019 two, 2002, made that leapreally from tech law to legal tech.
How does the loop platform work?
Loophole is a work management system, andthe problem that we solve is that what

(01:29):
we see in law firms is there's this biggap between, systems for the billing and
the time and systems for the documents.
The work product that rollsoff the production line.
But in the middle there's this wholespace, which is who's doing what and
what needs to happen next on my caseor my transaction, or my project.
And right now, a lot of that ismanaged in Excel trackers and

(01:53):
word tables and email threads.
Whilst every other industry hasadopted work management tooling, like
you can't run an engineering teamwithout a JIRA or one of those tools
or an IT team without ServiceNow.
Or a marketing functionwithout monday.com or Asana.
In legal, we have our word tables andour Excel trackers, and that's crazy.

(02:14):
So what we do is we take work outof those clunky manual trackers.
We put it into a real time collaborativesystem, wherever everyone can see what's
going on what needs to happen next.
And then increasingly,we started with the.
Hey, how can we get people toorganize their work to help
attorneys organize their work?
And now that we've organizedit, how can we automate it?

(02:36):
And we do that with bothdeterministic automation.
So that's the whole, ifthat then this workflows.
But we also do itincreasingly with AI and.
The reference you just made there toattorney plus agent Teams is a new
feature, really a new product that we'verolled out at the start of this year,
which is, we think AI is gonna go frombeing what it is now mostly, a chat bot.

(03:01):
You look at how it's being used in mostlaw firms, it's a reactive chat bot.
Already in engineeringand other industries, it's
becoming a proactive teammate.
So how do we apply everythingwe've learned about.
Attorney to attorney collaboration andproject management for the last five,
six years to a new paradigm where, yes,it's attorneys, working with attorneys,
but it's also attorneys workingwith agents alongside them as well.

(03:25):
That's the bet that we've made herereally in the start of 2026, that
is gonna be the new way of working.
How much human input and how muchautomation is necessary for a team
to optimize the loophole platform.
We are very much believers in human, inthe loop, attorney in the loop, we do not
think, and, for insurance reasons, forliability reasons, for all the reasons

(03:50):
which we both know, we don't see aworld right now where it's just agents.
But let's take the exampleof like contract review.
There's a growing numberof AI law firms launching.
And what these law firms tend to dois they'll intake a client request.
You need a bunch of contracts reviewed.
You send it out to the law firm.
AI does the first pass.

(04:11):
It triages it, and it says, okay,, eight of the 10 are aligned with your
playbook, but two of them look high risk.
So we're gonna assign thoseto a more senior lawyer.
The senior lawyer reviews itand they get back to you within.
In some cases, these guys are doing likea one hour SLA or a three hour SLA so
that's what the new entrants are doing.
We are effectivelydoing that with big law.

(04:33):
We're helping big law to adoptthose kinds of processes.
You have a first pass review byan AI agent who takes an email
or a teams message triages.
It does the first pass.
And then assigns it to the human reviewer.
I could do others data, subjectaccess requests litigation
filings and things like that.

(04:53):
But it's all about, how can AInot just sit there and wait for
you to ask it to do a thing?
How can it do a thing even whilstyou're in all day meetings?
It's done the first pass, it'smoved things along, and it's and
then we can take the next stepin whatever the legal process is.
How did you evolve loophole'sinitial mission to adapt to the

(05:14):
current needs of legal professionals?
We launched in the middle of thepandemic and no one was really using
collaboration tools in 2018 2019.
Do you remember the days of the.
You get on a conference call and there'sthe ping when Ari has joined the meeting.
And then sometimes somebody doesn'tannounced they've joined the
meeting and you don't know they'rethere and all of that fun stuff.

(05:35):
It's like a lifetime ago.
And that's actually when we launchedLoop was really a lifetime ago,
feels like before the pandemic.
And we evolved the product from itsinitial focus, which was just general
collaboration and communication, whichvery quickly got swallowed up into
the world of teams and Slack and Zoomand it wasn't differentiated enough.
So we effectively pivoted as a company tofocus on work management, legal project

(06:00):
management, and workflow automation,because that was the part of the product
that people were really engaging withand they were less interested in just
general communication type tooling.
So that's how we evolvedand really more of a pivot.
What's interesting, I think isjust in the last few months.
We've noticed the Harvey's and thelags starting to do some of what
we were doing five years ago withshared spaces and collaboration

(06:22):
portals and things like that.
Less so on the work management side.
But, I guess everything changes andeverything stays the same in some ways.
When I used to join those calls and Iremember those days where it'd be like.
Matt is joining the call.
Bob is joining the call and because youhad to record your name, I'd say something
like Ari Kaplan then all these peoplewould be, Susan has joined the call.

(06:48):
Ari like it.
People just start laughing.
And I thought, come onguys, this is super boring.
Let's let's ramp it up a little bit.
And there was always one guy whodidn't announce himself properly,
and you'd get half, you're 30 minutesinto the call and you're talking
about Dave as if he's not there.
And suddenly he chimes in withhe said, and be like, Dave, I
didn't know you were in the room.
You, you've been here this whole time.

(07:08):
Good times.
Why did you launch Vibe Code Law?
Vibe code.law.
, So I have a bad habitof buying domain names.
My wife criticizes me for this and right,rightly but it literally started with me.
I've been vibe codingfor a couple of years.
Like I love building stuff.
I do it in my day job.
But I also do it for thisfun weekend projects and.

(07:31):
I saw the Vibe Code, law Domainwas available, bought it, didn't
have a immediate plan for it.
And then got together with a fewfriends and we spent a couple
of weekends building some stuff.
And the reason for doing it really iswe felt like the conversation on Vibe
coding is polarized between, on theone hand, you've got people who say,
I could vibe Code Harvey in a weekend.

(07:52):
And there are the others whoare saying it's a waste of time.
You can't build anything.
You're not serious ifyou're vibe, coding stuff.
And neither of those twothings are true, in my opinion.
I think the right answeris somewhere in the middle.
So we trying to bring a bit of senseand clarity to the conversation.
It's also just great fun.
And we've just launched athing called Vibe Academy,
which is a place where legal.

(08:13):
Professionals can learn to build.
It's all totally free.
It's not a commercial thing.
It's genuinely like a hobby project,which we're running on the weekends.
But the reaction has been great.
And we've got, I think coming up for 40projects on there from around the world.
And I've gotta say the coolestone was from launching.
Within a week we had the Harvey ofMongolia had submitted to the platform.

(08:35):
I just love that it went fromidea to release to someone in Ulu.
Batar is sending in an appin the space of a week.
It's just the power ofLinkedIn and the internet,
what is the underlying technology forthem to vibe, code through your site?
They don't actually vibe codedirectly through the site.
We effectively provide a place for'em to share what they've built

(08:55):
and to learn how to build stuff.
We recommend a few different tools.
So for someone who's just gettingstarted with Vibe coding, we
usually recommend Google AI Studio.
Because it's free up to a pretty generouslimit and you can genuinely vibe code
an app in 30 seconds to a minute.
It's got that wow factorthat immediate dopamine hit.

(09:16):
I did this with a group of lawstudents a couple of weeks back on a.
Presentation and we vibe coded an app.
The TikTok of Englishcase law was the idea.
We made it in 30 seconds or a minute.
And that was AI Studio.
There are a bunch of others,repli lovable, they're
quite good for prototypes.
Claude Code is the king right now.
They built Claude Cowork with Claude Code.

(09:38):
I don't know if you wouldcall that vibe coding.
The founder says he doesn'twrite code manually anymore,
hasn't done since December.
So I feel like the termitself is broadening to
encompass virtually all coding
is vibe coding safe for legal work.
We recommend what we callresponsible vibe coding for lawyers.

(09:59):
We always recommend don't uploadconfidential or privileged data
into something you've just built.
It's easy to, get aheadof yourself a little bit.
And we're all guilty of this.
You type in a prompt, an app appearsand you're like, wow I'm amazing.
I know how to build stuff.
But there's such a huge gap betweensomething that you build in a couple
of minutes that looks great, andproduction grade, enterprise grade

(10:23):
software, and particularly in ourindustry, there's just too much risk.
And in the same way as wewouldn't recommend a client.
Vibe, draft a share purchase agreementon a big m and a transaction.
Probably not a good idea tovibe code, something that's
gonna process confidential data.
Reach out to a proper experiencedengineer who can help you take what you've

(10:43):
built and take it to the next level.
Something that's production.
Prior to launching Loophole, youwere a partner at a large law firm.
How does that experience informthe work that you do at Loophole?
I was a tech lawyer back in the day.
I definitely apply a lot of what Ilearned from working in the tech sector.

(11:04):
I probably had an advantage over otherlawyers who hadn't necessarily worked
with startups, but still there's sucha difference between advising a startup
and actually being in a startup.
And that's been quite the learning.
Process, quite the rollercoaster ride.
It's not for the faint of heartfounding a startup, I will say that.

(11:24):
But we've had some greattimes and I love it.
But how does it inform it?
I think there's partly alsosome stuff I've had to unlearn.
As an attorney, you are sofocused on finding problems.
The racing track analogy is,we're focused on the barriers
around the edge of the track.
We're not focused on the road.
Sometimes.
That's probably a bit ofan unfair generalization.

(11:44):
And as a startup founder, you'vegotta see opportunities everywhere.
You can't just focus on the risks.
And you've gotta balance thosethings a little bit more.
But I'd recommend it toanyone, any attorney listening.
It's not for the faint of heart, butit's an incredible journey and the
barriers to entry to founder companythese days are lower than ever as well.
Why do you think so many lawyers seem tobe starting legal technology ventures?

(12:11):
It is a couple of things.
Obviously we are very close to theproblem space and there's still plenty
of problem space out there that'sunsolved, despite all the press releases
and the funding rounds and everything.
The fact that there are may be morelegal tech companies than law firms
these days there are still plentyof problems to solve and lawyers
experience those every single day.
When they see them unsolved, theywant to do something about it.

(12:33):
There's a sense perhaps of fomo.
You see a junior associate from a westcoast law firm found an $11 billion legal
tech company and you think, gosh, maybeI should do something there as well.
So there's a bit of that.
And then, to the discussion justnow on vibe coding, we are at a
time where a partner or a foundercan vibe, code a prototype.
Again, not a productionapp, but a prototype.

(12:54):
Get it in the hands of potentialusers say, would you use this?
Maybe get it in the hands ofinvestors and get some seed funding
and start to build something.
The days of look, it's gonna cost you250 K to get a prototype behind us.
You can move so much fasternow and I guess that's gonna be
appealing for a lot of attorneys.
Turn founders.
How do you see thepractice of law evolving?

(13:18):
The revolution's not gonnabe evenly distributed.
Maybe I'll answer it for big law.
We will see a growingproductization of legal work.
It's already starting to happen,but we're gonna end up with what I
call a multi-channel law firm where.
Some categories of work that used tobe done in a very artisan manual way.

(13:38):
Hourly rates, it's justnot gonna happen anymore.
Either the clients are not goingto outsource it, or if they do,
they're going to want fixed fee.
Fixed SLA turnaround time and outcomes.
They're not gonna be paying for inputs.
They want to know.
What the outcome of that is going to be.
So a bunch of work will flowdown that productized channel.
Which is scary for some law firms because,do you cannibalize your own practice

(14:03):
groups and target that productized work?
'cause you know where it's going.
The whole iPhone cannibalize the.
iPod.
It does seem to be where it's going.
That's the product size channel.
But then there will always be roomfor the bet the company stuff.
So the run the company stuff goes downthe product size channel, the bet the
company stuff probably isn't gonna changethat much, and probably the hourly rates
for that are gonna go up even higher.

(14:26):
This is Ari Kaplan speaking with MattPollins, the co-founder and chief product
Officer at Loophole, a work managementplatform for attorneys and agent teams.
He's also the co-founder of Vibe Code Law,an open community platform launch this
year for legal professionals to learn howto build apps and share what they've made.
Matt, thank you so very much.

(14:48):
Thanks again.
Thank you for listening to theReinventing Professionals Podcast.
Visit reinventing professionals.com orari kaplan advisors.com to learn more.