May 5, 2025 • 22 mins
I spoke with Uwais Iqbal, the Founder of Simplexico, a UK-based AI consulting firm helping legal teams with AI adoption. We discussed the most common use cases for generative AI, how legal professionals can maximize their use of generative AI, whether law firms should build or buy generative AI applications, and how the use of AI in law firms is evolving.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(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 Uwais Iqbal, the founder
(00:24):
of Simplexico, a UK based AIconsulting firm helping law firms
and legal teams with AI adoption.
Hi Uwais, how are you?
Hey Ari, thanks for having me.
I'm doing good, thanks.
I'm looking forward to this conversation,so tell us about your background
and the genesis of Simplexico.
I'm not a lawyer.
(00:44):
Never went to law school, neverpracticed law, never taken, Any legal
certification exams or anything like that.
My interest in the legal field comesfrom the point of view of a technologist,
I've spent close to a decade workingas a data scientist, a machine learning
engineer, an NLP practitioner, AIpractitioner, Pick your term of choice.
Effectively, a technologist who's beenworking and designing and developing
(01:06):
AI tools in the legal sector.
I've held roles across companieslike Eigen Technologies, Thought
River, as well as Stint within theInnovation Lab at Thomson Reuters.
I've seen , a spectrum of approachesof technology with AI in particular
within legal from the startup.
World as well as the incumbent, andwhat interested me was applying AI
(01:27):
in legal, because it's a very nichedomain and there's a lot of the
constraints you have around legal aremore complex and you have to think about
solving problems in interesting ways.
As I was going through my journey ofworking through these companies, I
realized I spent time, writing anddeveloping the same models algorithms
and code for all of these companies.
And everyone was taking that technologylayer and applying it to different
(01:49):
application use cases . And therewas a massive layer of marketing
stacked on top saying, look at us,we're using AI, we're doing something
so innovative come talk to us.
So it always struck me as very interesting
.A lot of the legal techs.
solutions are actually the samefirst hand experience of building
these things at different companies.
It struck a chord with me to thinkwhether or not there's a better way
(02:10):
to deliver AI to the legal industry.
So I had my circumstances aligned forme to jump ship and try starting my
own business as a bit of an experimentand then that was the synthesis and
the catalyst for starting SuplexicalThe idea was to do something in through
a services based model, as opposed toa product SAS delivery mechanism for
delivering technology and AI into legal.
(02:30):
We decided not to takeon any venture funding.
So we're completely bootstrapped andwe've been running for about two and a
half years . I started the company inAugust of 22 and the timing was really
fortunate because it was only a coupleof months later that the whole gen AI.
Flywheel kicked into place and everyonegot interested in chat GPT became
a thing and then legal AI actuallybecame a category in and of its own.
(02:51):
Over the past two and a half years,we focused on a number of key areas,
but it's becoming clear where thekey pillars are for us as a company.
The first is around AIeducation and advisory.
How do we help educate the legalmarket as well as advise firms on
how to approach the technology,how to be strategic with AI.
Second area is around usecase discovery and design.
That's focusing particularly onspecific use cases and how to
(03:14):
structure them, create ROI aroundthem, turn them into business cases
and promote them within organizations.
And the third area, which is, thearea I'm most excited about is bespoke
implementation of AI solutions forvery specific and issues cases.
We've ended up working with fanslike linkators, but in bed shown here
we've done some work for the tenancydeposit scheme here in the UK, and
we're working with a number of clients.
(03:34):
We're still very much inthat phase of getting our.
Product market fit exists on theproduct side, but the service market
fit, we're in the space of gettingour service market fit, right.
And tuning that in with customersbefore we figure out how to
scale and grow the business.
What are the most common use casesthat you're seeing for generative AI?
It's always good to take a step backand start from first principles of
(03:59):
how do we actually think about AI?
How do we define it?
And then build on top of thatto think about use cases.
At Simplexical, we like to use thedefinition of what we think about AI
as a technique where we want to takewhat humans can do with their minds
when it comes to problem solving anddecision making and replicate or mimic
that through the use of machines.
If you have this definition in yourmind, AI is about taking what humans
(04:19):
do through cognitive skills andintelligence and leveraging machines
to mimic or replicate aspects of that.
That's AI in a broad sense.
If you narrow that down and focuson legal specifically, then you just
replace human with legal professional.
What are the problems and decisionslegal professionals make in the
day to day practice of law, wherewe want to try and explore how
(04:41):
machines can mimic some of those.
solving some of those problems orso quick making those decisions.
So it's this idea of lawyersalready doing what they're doing.
They're excellent at it, but arethere areas where we can try and
replicate what they do throughthe use of machines and where they
solve problems and make decisions?
Can machines be offered as an alternativeor to some degree automate entirely?
(05:04):
If you have that in mind, we've beenworking with, law firms and legal
customers to actually Try and identifythe key areas of activity that legal
professionals and lawyers do, wherethere's an overlap with what machines
can do really well, where they are quite,they excel at replicating these tasks.
And we've come up with whatwe call the legal AI actions.
These are things like extractinginformation, labeling information.
(05:27):
Drafting comparing, organizing,finding, interrogating, summarizing,
translating, transcribing.
For us, it's really useful to takethe conversation away from very vague
and broad buzzwords like AI, GenAI, LLMs, and instead focus on the
concrete realities of What lawyers areactually doing on a day to day basis.
So take the conversation away fromthis gen AI use case, which is very
(05:49):
vague and make it very specific around.
Real estate agreements.
I want to do extraction of terms tounderstand what the portfolio looks
like, then that's a much more concreteway of approaching the technology.
We like to approach itfrom this perspective of.
Being very clear about the particular useof the technology and for us, it's been
interesting as we worked with customers.
(06:10):
We've uncovered different types of usecases within firms and within teams.
There are internal facing use casesand external client facing use cases.
The internal use cases are typicallymore transformation focused where
there's an existing workflow.
Lawyers are already doing thisworkflow or this service delivery.
And the goal is to try and embed AI intothat practice to drive efficiency, to
(06:34):
remove bottlenecks, to increase scaleand help the bottom line effectively so
you can get some return on investment.
Those are more transformation use cases.
How do you take an existing legalworkflow and create it to something new?
Then on the client facing side, there'smore innovation focused use cases.
These are blue skies thinking around howcan we leverage the firm's knowledge,
capability and expertise to standup use cases where we can leverage
(06:58):
the technology to serve clients inways we couldn't previously serve.
That's the most excitingarea for these new.
forward looking use cases.
Then within firms, there's, use casesaround the business of law, which
are HR, sales, marketing, finance.
There's use cases around legal operations.
There's a lot of data around billing,matter management, resource estimation.
All of that can be fed into AI systems.
(07:20):
And I think the real focuswith GNI has particularly been
around the practice of law.
If you're practicing law, whether you'redoing legal research, whether you're
doing due diligence, whether you'redoing contract review and analysis,
how GNI can be applied in those areas.
So what's the most common use cases?
It depends on what the activitiesof the lawyers actually are.
From a high level, there's use casesaround from a legal context, there's
(07:42):
use cases around legal research,there's use cases around contract
management and review, there's usecases around due diligence, and
there's, use cases around mattermanagement and predictive analytics.
And the sum total of the landscape ofuse cases, it's still quite mature,
I think a lot of work needs to bedone to actually help map out and
(08:02):
create the language and thinkingaround these cases in a mature way.
How can legal professionals learnmore about how to use and maximize?
Generative AI.
It starts with being curious and open.
I'm having a go with the technology.
So signing up to some of the freetools which are available, whether
that's chat GPT, Gemini from Googleor Claude playing around with these
(08:25):
tools, getting a feel for them.
And seeing what they can do andmore importantly what they can't
do and within your organization.
If there's tools available then gettinginvolved with innovation initiatives
or with pilot projects so you can testthese tools and determine whether they're
fit for purpose or not fit for purpose.
The thing which always interested mefrom from working as a technologist
(08:47):
in the legal domain is that thearbiters or the decision makers
for success in legal are not thetechnologists, but the domain experts.
The lawyers and the legal professionals,have the final say over whether technology
is actually useful not necessarily thetechnologists or the investors themselves.
It's actually the legalprofessionals and the main experts.
So we need legal domain experts to lead.
(09:09):
Be vocal challenge the narrative andtry and push technology vendors, push
technologists to actually think how toimprove technology within the technology,
AI, gen AI the implementation andapplication within the legal industry for
more practical resources as in Plexiglaswe just put together a free email course
last month, which is a free seven dayemail course directly to your inbox.
(09:32):
If anyone's interested ingetting a bit more of a guided
Walkthrough of the basics of AI.
You can find that on our websiteover the past two and a half years.
We've trained thousands of legalprofessionals and we've condensed all of
that know how and expertise of what thewhat people in legal needs to know about
AI into a short seven day email course.
If anyone's interested inan actual resource, you can
(09:52):
head over to simplexical.
ai and access it there.
What's the best way for professionalswith existing access to generative
AI tools to identify the best way touse it in their particular workflows?
We've created a design toolkit andworkshop we run with customers to help
them around use case discovery and usecase identification within practice teams.
(10:13):
The starting point is getting clearon what AI is as a technology.
We want to mimic what legalprofessionals can do when it comes to
problem solving and decision making.
And then start with the tasks they arealready performing within their workflows.
Are you doing extractionwithin your workflows?
Are you doing any labelingwithin your workflows?
Are you doing any comparison ofclauses within your workflows?
(10:36):
Are you doing any interrogationof documents or evidence
within your workflows?
If you have the actions as your backbone,exploring your current workflows
and trying to identify where any ofthose actions are being performed by
humans and legal professionals today,Those are candidate opportunities
for where AI can be applied.
Because if there's lots of extractiontaking place within a due diligence
(10:57):
workflow, for example, that's acandidate opportunity for an AI
system to be introduced to do someof that due diligence instead of
relying completely on a machine.
From a high level, , the stepsinvolved are getting clear on the
actions that AI is really good at.
What we call the legal AI actions,mapping out the current workflow
for how lawyers do things in a verysystematic and process oriented way,
(11:20):
identifying any actions where AI canbe applied, and then redesigning a
workflow where machines and humansare embedded within service delivery.
It's becoming a bit of an art andthe conversation around use cases
varies . A use case for an individualmight be something as broad as
summarization, and for another person,a use case might be very specific.
(11:42):
For example, as an adjudicator,how do I use AI to interrogate or
interrogate evidentiary documentsto minimize time to decision?
It can range from a very finegrained use case to a very broad
or course grained use case.
Getting as precise as possible makesit much more effective to actually
measure success and think about adoptionand transformation of the technology.
(12:05):
Should law firms build or buygenerative AI applications?
This is a polarizing question.
At the start of the year, I rana series of listening interviews
with law firm innovation leaders.
So I spoke to 14 law firm innovationleaders from across the globe.
And it was very interesting because thisis one of the questions I posed to them.
(12:28):
And I was surprised by howpolarizing this question is.
Some innovation leaders wereCompletely against the build approach
I remember one innovation leader.
Said anybody who builds is stupid.
And he's a fool and law firmsshould just focus on why.
Then I spoke to other law firm innovationleaders who were of the mindset that
law firms should just be building.
It doesn't make sense to buy technology.
(12:49):
We need to leverage our data.
There's a spectrum of, positions orapproaches . It's a bit of a nuanced
question because there's some subtletybased on the specifics of a firm or what
particular strategy a firm has in place.
On one end of the spectrum, you have.
Microwaves.
These are multi purpose tools.
(13:09):
They can do lots of things, but theycan do them to a very low quality or a
very low degree of quality and fidelity.
You could warm some food.
After the microwave, you havesomething like a pizza oven.
A focused application . You canonly cook pizzas in it, but the
pizzas are always going to be betterthan any other thing you can make.
At the other end of the spectrum, you havepersonal chefs where it's a completely
(13:32):
customized approach where you're in thekitchen, using high quality ingredients
and creating things according to arecipe, according to what particular
taste profile the customers might have.
There's an analogy into the spectrumfor different use cases, . There's
a buy opportunity for microwavesand pizza ovens, and then there's
a build opportunity for use caseswhich require a personal chef.
(13:54):
It's not as easy as saying we're justgoing to buy a microwave and that's
all we're going to have in our kitchen.
Because a lot of legal work, it'sless like cooking a jacket potato
and it's more like making a souffle.
It's very high fidelity.
It's very high quality Michelin star.
It's almost like Michelinstar preparation.
You need that finesseand quality around it.
Firms should take a position wherethey're buying microwaves and
(14:17):
pizza ovens where it makes sense.
Then they're using personalchefs or bringing in AI expertise
and talent where they want to
.strategically focus on use cases which can drive value for the firm and
differentiate them with the technologywhere it makes sense to do and we're
seeing this increasingly with AI thatmore and more firms are adopting a hybrid
strategy buying some solutions but alsobuilding to tinker and learn about the
(14:40):
technology or building and partneringwith providers to develop strategic use
cases for the firms in at Simplexicalwe're thinking about how we can position
ourselves almost as the personal AIchefs for law firms who don't have access
to development teams and AI talent.
How can we partner and be a fractionalAI team to help law firms develop
(15:01):
solutions and focus on those, buildapproaches for those types of use cases.
What are the advantages anddisadvantages of each approach?
On the build side, the pros have alwaysbeen there from the software side.
If you're buying , you get access toa product which has been informed by
(15:21):
other users, and you get access toengineering talent that can build that
product, you don't have to worry aboutmaintenance and updates, you just
worry about consuming or using theproduct, you don't really worry about
designing, offloaded on the vendor side.
But you always have to deal withthat layer of marketing, or that
layer from product marketing andsales before you can actually
(15:42):
benefit from the engineering talent.
Some of the disadvantages of buyingthere's vendor lock in, you're stuck
into using a particular vendor.
So if you want to move yourdata around or move your models
around, there's challenges.
You're at the mercy of the marketand the pricing of the market.
So with the technology beingvery I'm nascent with Jenny.
I lots of the Jenny.
I microbes and pizzaovens are very expensive.
(16:04):
It prices out a lot of mid market firms.
Only elite law firms can go after them.
This is part of the dynamic of howlegal tech is funded in terms of these.
solutions which couldbe bought by law firms.
They're backed by VC funding.
And that means founders tend tofocus on problems which can give
them enough exponential scale sothat investors can get return on
(16:27):
their money in a short enough windowfor the investors to be happy.
So these products tend to focuson large scalable use cases.
There's lots of AI startups focusingon things like legal research, due
diligence, contract review, becausethere's that size of market for investors
to be confident going after, but thenthere's a whole, long tail of problems
and issues in legal tech that stillneed technology to be applied to.
(16:51):
So if you're buying, you're only goingto be stuck in that top realm of what's
available on the market on the build side.
There's advantages and disadvantages.
Some of the disadvantages youneed access to expert AI talent.
It's going to take a bitmore time to actually build
and learn to get through it.
Some of the advantages ofbuilding, in an AI world is that
the kitchen has become available.
(17:13):
With.
Cloud enterprise platforms like AzureAI Foundry and things like this.
Every law firm, every organizationhas access to a state of the art
kitchen where they can get access tothe same models that are powering the
latest AI applications like ChatGPT
.Even deep seeks become available on Azure AI Foundry so people can access the
(17:33):
models and then they can actually buildon the application layer and don't have to
worry so much about the technology layer.
On the build side, there's a biggeradvantage for law firms as well from a
compliance standpoint, because there'smuch more visibility and transparency
in terms of how data is being processed,where data is being stored, and from an
enterprise view, if it's built in theright way, the cloud architecture is
(17:55):
set up and designed in the right way.
It can mean that data can stay within thefirm and still access the technologies.
Whereas if you're working with avendor there's going to be a third
party involved in processing the data.
How do you see the use ofAI in law firms evolving?
There's a few trends which woulddictate how AI is used within law firms.
(18:19):
AI has been targeted to the top law firms.
What we'll see over the next few yearsis that coming down market to target mid
and smaller law firms So they can alsoleverage the technology And with trends
around agentic AI making headlines for2025 Law firms will need to think more
carefully about AI, as opposed to justbeing a one time purchase as a status
(18:42):
symbol to show we're innovative, toactually think very deeply about their
business model and very deeply abouthow they can integrate AI within their
operations as well as the practiceand delivery of legal services.
From the perspective of, Computers faxmachines or printers being introduced.
People would buy theminitially to demonstrate that
they're ahead of the market.
But at some point, the entire firm hadto get onto a journey, a transformation
(19:06):
journey of adopting PCs laptops computershardware and document management
systems to be effective as a law firm.
It was embedded into the practice.
We're going to see asimilar thing with AI.
except that law firms are in the bestposition to make the most use of AI . AI
is fundamentally driven by what datayou have access to or what ingredients
(19:28):
you can access for the personal chef.
Law firms already have the kitchen.
From the cloud enterprise provider.
They've got access to the ingredients.
They've got access to decades of matter,precedents documentation knowledge
content, all available within the firm.
The key ingredient they'remissing is the talent piece.
If law firms are able to figure out, Howto bring in AI talent into that ecosystem
(19:51):
of integrating technology and AI intotheir practice And into legal services.
Things will be very exciting we'll seea shift from the knowledge management
perspective where it becomes less aboutcurating data to benefit humans so
that humans can leverage it to curatingknowledge and expertise into data so
that machines can also leverage it.
(20:14):
We like to call this intelligenceengineering, where it will go from just
managing and taking knowledge of theheads of people to actually curating
intelligence that can feed machines,which can then subsequently feed humans.
With the trend of agentic AI.
What we're going to see is the riseof an agent agnostic law firm or agent
agnostic enterprises and organizations.
(20:36):
Typically when we speak about agency,we've only really considered humans
to have agency or to be able tomake decisions and complete tasks.
Now with agents, machines are also goingto have some agency to have autonomy,
to make decisions and complete tasks.
What we're going to see potentially inthe future is organizations and all firms
(20:56):
where work comes in and is orchestrated.
Part of that work is offloadedto a human and part of that
work is offloaded to a machine.
It becomes almost as thoughthat work is agent agnostic.
Whether a machine completes that workor whether a human completes that
work doesn't make a difference aslong as the work product gets done.
(21:16):
So from a future, Looking perspective.
It's really important for law firms to goback to those central questions around.
How do you make sure your datais in a good space internally?
How do you make sure you havethe right infrastructure in
place to enable transformation?
And how do you make sure that from astrategic perspective, AI is part of
the charter of the firm's growth anddevelopment, as opposed to just being.
(21:41):
Something which demonstrates movementon the innovation theater stage.
It's just emotion without reallythinking about it in a very deep way.
This is Ari Kaplan speaking with uesIal, the founder of Plco, a UK based
AI consulting firm, helping law firmsand legal teams with AI adoption.
(22:01):
Oise, wonderful conversation.
Thank you so very much.
Thanks to have Mary.
Thank you for listening to theReinventing Professionals podcast.
Visit ReinventingProfessionals.
com or AriKaplanAdvisors.
com to learn more.
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 Uwais Iqbal, the founder
(00:24):
of Simplexico, a UK based AIconsulting firm helping law firms
and legal teams with AI adoption.
Hi Uwais, how are you?
Hey Ari, thanks for having me.
I'm doing good, thanks.
I'm looking forward to this conversation,so tell us about your background
and the genesis of Simplexico.
I'm not a lawyer.
(00:44):
Never went to law school, neverpracticed law, never taken, Any legal
certification exams or anything like that.
My interest in the legal field comesfrom the point of view of a technologist,
I've spent close to a decade workingas a data scientist, a machine learning
engineer, an NLP practitioner, AIpractitioner, Pick your term of choice.
Effectively, a technologist who's beenworking and designing and developing
(01:06):
AI tools in the legal sector.
I've held roles across companieslike Eigen Technologies, Thought
River, as well as Stint within theInnovation Lab at Thomson Reuters.
I've seen , a spectrum of approachesof technology with AI in particular
within legal from the startup.
World as well as the incumbent, andwhat interested me was applying AI
(01:27):
in legal, because it's a very nichedomain and there's a lot of the
constraints you have around legal aremore complex and you have to think about
solving problems in interesting ways.
As I was going through my journey ofworking through these companies, I
realized I spent time, writing anddeveloping the same models algorithms
and code for all of these companies.
And everyone was taking that technologylayer and applying it to different
(01:49):
application use cases . And therewas a massive layer of marketing
stacked on top saying, look at us,we're using AI, we're doing something
so innovative come talk to us.
So it always struck me as very interesting
.A lot of the legal techs.
solutions are actually the samefirst hand experience of building
these things at different companies.
It struck a chord with me to thinkwhether or not there's a better way
(02:10):
to deliver AI to the legal industry.
So I had my circumstances aligned forme to jump ship and try starting my
own business as a bit of an experimentand then that was the synthesis and
the catalyst for starting SuplexicalThe idea was to do something in through
a services based model, as opposed toa product SAS delivery mechanism for
delivering technology and AI into legal.
(02:30):
We decided not to takeon any venture funding.
So we're completely bootstrapped andwe've been running for about two and a
half years . I started the company inAugust of 22 and the timing was really
fortunate because it was only a coupleof months later that the whole gen AI.
Flywheel kicked into place and everyonegot interested in chat GPT became
a thing and then legal AI actuallybecame a category in and of its own.
(02:51):
Over the past two and a half years,we focused on a number of key areas,
but it's becoming clear where thekey pillars are for us as a company.
The first is around AIeducation and advisory.
How do we help educate the legalmarket as well as advise firms on
how to approach the technology,how to be strategic with AI.
Second area is around usecase discovery and design.
That's focusing particularly onspecific use cases and how to
(03:14):
structure them, create ROI aroundthem, turn them into business cases
and promote them within organizations.
And the third area, which is, thearea I'm most excited about is bespoke
implementation of AI solutions forvery specific and issues cases.
We've ended up working with fanslike linkators, but in bed shown here
we've done some work for the tenancydeposit scheme here in the UK, and
we're working with a number of clients.
(03:34):
We're still very much inthat phase of getting our.
Product market fit exists on theproduct side, but the service market
fit, we're in the space of gettingour service market fit, right.
And tuning that in with customersbefore we figure out how to
scale and grow the business.
What are the most common use casesthat you're seeing for generative AI?
It's always good to take a step backand start from first principles of
(03:59):
how do we actually think about AI?
How do we define it?
And then build on top of thatto think about use cases.
At Simplexical, we like to use thedefinition of what we think about AI
as a technique where we want to takewhat humans can do with their minds
when it comes to problem solving anddecision making and replicate or mimic
that through the use of machines.
If you have this definition in yourmind, AI is about taking what humans
(04:19):
do through cognitive skills andintelligence and leveraging machines
to mimic or replicate aspects of that.
That's AI in a broad sense.
If you narrow that down and focuson legal specifically, then you just
replace human with legal professional.
What are the problems and decisionslegal professionals make in the
day to day practice of law, wherewe want to try and explore how
(04:41):
machines can mimic some of those.
solving some of those problems orso quick making those decisions.
So it's this idea of lawyersalready doing what they're doing.
They're excellent at it, but arethere areas where we can try and
replicate what they do throughthe use of machines and where they
solve problems and make decisions?
Can machines be offered as an alternativeor to some degree automate entirely?
(05:04):
If you have that in mind, we've beenworking with, law firms and legal
customers to actually Try and identifythe key areas of activity that legal
professionals and lawyers do, wherethere's an overlap with what machines
can do really well, where they are quite,they excel at replicating these tasks.
And we've come up with whatwe call the legal AI actions.
These are things like extractinginformation, labeling information.
(05:27):
Drafting comparing, organizing,finding, interrogating, summarizing,
translating, transcribing.
For us, it's really useful to takethe conversation away from very vague
and broad buzzwords like AI, GenAI, LLMs, and instead focus on the
concrete realities of What lawyers areactually doing on a day to day basis.
So take the conversation away fromthis gen AI use case, which is very
(05:49):
vague and make it very specific around.
Real estate agreements.
I want to do extraction of terms tounderstand what the portfolio looks
like, then that's a much more concreteway of approaching the technology.
We like to approach itfrom this perspective of.
Being very clear about the particular useof the technology and for us, it's been
interesting as we worked with customers.
(06:10):
We've uncovered different types of usecases within firms and within teams.
There are internal facing use casesand external client facing use cases.
The internal use cases are typicallymore transformation focused where
there's an existing workflow.
Lawyers are already doing thisworkflow or this service delivery.
And the goal is to try and embed AI intothat practice to drive efficiency, to
(06:34):
remove bottlenecks, to increase scaleand help the bottom line effectively so
you can get some return on investment.
Those are more transformation use cases.
How do you take an existing legalworkflow and create it to something new?
Then on the client facing side, there'smore innovation focused use cases.
These are blue skies thinking around howcan we leverage the firm's knowledge,
capability and expertise to standup use cases where we can leverage
(06:58):
the technology to serve clients inways we couldn't previously serve.
That's the most excitingarea for these new.
forward looking use cases.
Then within firms, there's, use casesaround the business of law, which
are HR, sales, marketing, finance.
There's use cases around legal operations.
There's a lot of data around billing,matter management, resource estimation.
All of that can be fed into AI systems.
(07:20):
And I think the real focuswith GNI has particularly been
around the practice of law.
If you're practicing law, whether you'redoing legal research, whether you're
doing due diligence, whether you'redoing contract review and analysis,
how GNI can be applied in those areas.
So what's the most common use cases?
It depends on what the activitiesof the lawyers actually are.
From a high level, there's use casesaround from a legal context, there's
(07:42):
use cases around legal research,there's use cases around contract
management and review, there's usecases around due diligence, and
there's, use cases around mattermanagement and predictive analytics.
And the sum total of the landscape ofuse cases, it's still quite mature,
I think a lot of work needs to bedone to actually help map out and
(08:02):
create the language and thinkingaround these cases in a mature way.
How can legal professionals learnmore about how to use and maximize?
Generative AI.
It starts with being curious and open.
I'm having a go with the technology.
So signing up to some of the freetools which are available, whether
that's chat GPT, Gemini from Googleor Claude playing around with these
(08:25):
tools, getting a feel for them.
And seeing what they can do andmore importantly what they can't
do and within your organization.
If there's tools available then gettinginvolved with innovation initiatives
or with pilot projects so you can testthese tools and determine whether they're
fit for purpose or not fit for purpose.
The thing which always interested mefrom from working as a technologist
(08:47):
in the legal domain is that thearbiters or the decision makers
for success in legal are not thetechnologists, but the domain experts.
The lawyers and the legal professionals,have the final say over whether technology
is actually useful not necessarily thetechnologists or the investors themselves.
It's actually the legalprofessionals and the main experts.
So we need legal domain experts to lead.
(09:09):
Be vocal challenge the narrative andtry and push technology vendors, push
technologists to actually think how toimprove technology within the technology,
AI, gen AI the implementation andapplication within the legal industry for
more practical resources as in Plexiglaswe just put together a free email course
last month, which is a free seven dayemail course directly to your inbox.
(09:32):
If anyone's interested ingetting a bit more of a guided
Walkthrough of the basics of AI.
You can find that on our websiteover the past two and a half years.
We've trained thousands of legalprofessionals and we've condensed all of
that know how and expertise of what thewhat people in legal needs to know about
AI into a short seven day email course.
If anyone's interested inan actual resource, you can
(09:52):
head over to simplexical.
ai and access it there.
What's the best way for professionalswith existing access to generative
AI tools to identify the best way touse it in their particular workflows?
We've created a design toolkit andworkshop we run with customers to help
them around use case discovery and usecase identification within practice teams.
(10:13):
The starting point is getting clearon what AI is as a technology.
We want to mimic what legalprofessionals can do when it comes to
problem solving and decision making.
And then start with the tasks they arealready performing within their workflows.
Are you doing extractionwithin your workflows?
Are you doing any labelingwithin your workflows?
Are you doing any comparison ofclauses within your workflows?
(10:36):
Are you doing any interrogationof documents or evidence
within your workflows?
If you have the actions as your backbone,exploring your current workflows
and trying to identify where any ofthose actions are being performed by
humans and legal professionals today,Those are candidate opportunities
for where AI can be applied.
Because if there's lots of extractiontaking place within a due diligence
(10:57):
workflow, for example, that's acandidate opportunity for an AI
system to be introduced to do someof that due diligence instead of
relying completely on a machine.
From a high level, , the stepsinvolved are getting clear on the
actions that AI is really good at.
What we call the legal AI actions,mapping out the current workflow
for how lawyers do things in a verysystematic and process oriented way,
(11:20):
identifying any actions where AI canbe applied, and then redesigning a
workflow where machines and humansare embedded within service delivery.
It's becoming a bit of an art andthe conversation around use cases
varies . A use case for an individualmight be something as broad as
summarization, and for another person,a use case might be very specific.
(11:42):
For example, as an adjudicator,how do I use AI to interrogate or
interrogate evidentiary documentsto minimize time to decision?
It can range from a very finegrained use case to a very broad
or course grained use case.
Getting as precise as possible makesit much more effective to actually
measure success and think about adoptionand transformation of the technology.
(12:05):
Should law firms build or buygenerative AI applications?
This is a polarizing question.
At the start of the year, I rana series of listening interviews
with law firm innovation leaders.
So I spoke to 14 law firm innovationleaders from across the globe.
And it was very interesting because thisis one of the questions I posed to them.
(12:28):
And I was surprised by howpolarizing this question is.
Some innovation leaders wereCompletely against the build approach
I remember one innovation leader.
Said anybody who builds is stupid.
And he's a fool and law firmsshould just focus on why.
Then I spoke to other law firm innovationleaders who were of the mindset that
law firms should just be building.
It doesn't make sense to buy technology.
(12:49):
We need to leverage our data.
There's a spectrum of, positions orapproaches . It's a bit of a nuanced
question because there's some subtletybased on the specifics of a firm or what
particular strategy a firm has in place.
On one end of the spectrum, you have.
Microwaves.
These are multi purpose tools.
(13:09):
They can do lots of things, but theycan do them to a very low quality or a
very low degree of quality and fidelity.
You could warm some food.
After the microwave, you havesomething like a pizza oven.
A focused application . You canonly cook pizzas in it, but the
pizzas are always going to be betterthan any other thing you can make.
At the other end of the spectrum, you havepersonal chefs where it's a completely
(13:32):
customized approach where you're in thekitchen, using high quality ingredients
and creating things according to arecipe, according to what particular
taste profile the customers might have.
There's an analogy into the spectrumfor different use cases, . There's
a buy opportunity for microwavesand pizza ovens, and then there's
a build opportunity for use caseswhich require a personal chef.
(13:54):
It's not as easy as saying we're justgoing to buy a microwave and that's
all we're going to have in our kitchen.
Because a lot of legal work, it'sless like cooking a jacket potato
and it's more like making a souffle.
It's very high fidelity.
It's very high quality Michelin star.
It's almost like Michelinstar preparation.
You need that finesseand quality around it.
Firms should take a position wherethey're buying microwaves and
(14:17):
pizza ovens where it makes sense.
Then they're using personalchefs or bringing in AI expertise
and talent where they want to
.strategically focus on use cases which can drive value for the firm and
differentiate them with the technologywhere it makes sense to do and we're
seeing this increasingly with AI thatmore and more firms are adopting a hybrid
strategy buying some solutions but alsobuilding to tinker and learn about the
(14:40):
technology or building and partneringwith providers to develop strategic use
cases for the firms in at Simplexicalwe're thinking about how we can position
ourselves almost as the personal AIchefs for law firms who don't have access
to development teams and AI talent.
How can we partner and be a fractionalAI team to help law firms develop
(15:01):
solutions and focus on those, buildapproaches for those types of use cases.
What are the advantages anddisadvantages of each approach?
On the build side, the pros have alwaysbeen there from the software side.
If you're buying , you get access toa product which has been informed by
(15:21):
other users, and you get access toengineering talent that can build that
product, you don't have to worry aboutmaintenance and updates, you just
worry about consuming or using theproduct, you don't really worry about
designing, offloaded on the vendor side.
But you always have to deal withthat layer of marketing, or that
layer from product marketing andsales before you can actually
(15:42):
benefit from the engineering talent.
Some of the disadvantages of buyingthere's vendor lock in, you're stuck
into using a particular vendor.
So if you want to move yourdata around or move your models
around, there's challenges.
You're at the mercy of the marketand the pricing of the market.
So with the technology beingvery I'm nascent with Jenny.
I lots of the Jenny.
I microbes and pizzaovens are very expensive.
(16:04):
It prices out a lot of mid market firms.
Only elite law firms can go after them.
This is part of the dynamic of howlegal tech is funded in terms of these.
solutions which couldbe bought by law firms.
They're backed by VC funding.
And that means founders tend tofocus on problems which can give
them enough exponential scale sothat investors can get return on
(16:27):
their money in a short enough windowfor the investors to be happy.
So these products tend to focuson large scalable use cases.
There's lots of AI startups focusingon things like legal research, due
diligence, contract review, becausethere's that size of market for investors
to be confident going after, but thenthere's a whole, long tail of problems
and issues in legal tech that stillneed technology to be applied to.
(16:51):
So if you're buying, you're only goingto be stuck in that top realm of what's
available on the market on the build side.
There's advantages and disadvantages.
Some of the disadvantages youneed access to expert AI talent.
It's going to take a bitmore time to actually build
and learn to get through it.
Some of the advantages ofbuilding, in an AI world is that
the kitchen has become available.
(17:13):
With.
Cloud enterprise platforms like AzureAI Foundry and things like this.
Every law firm, every organizationhas access to a state of the art
kitchen where they can get access tothe same models that are powering the
latest AI applications like ChatGPT
.Even deep seeks become available on Azure AI Foundry so people can access the
(17:33):
models and then they can actually buildon the application layer and don't have to
worry so much about the technology layer.
On the build side, there's a biggeradvantage for law firms as well from a
compliance standpoint, because there'smuch more visibility and transparency
in terms of how data is being processed,where data is being stored, and from an
enterprise view, if it's built in theright way, the cloud architecture is
(17:55):
set up and designed in the right way.
It can mean that data can stay within thefirm and still access the technologies.
Whereas if you're working with avendor there's going to be a third
party involved in processing the data.
How do you see the use ofAI in law firms evolving?
There's a few trends which woulddictate how AI is used within law firms.
(18:19):
AI has been targeted to the top law firms.
What we'll see over the next few yearsis that coming down market to target mid
and smaller law firms So they can alsoleverage the technology And with trends
around agentic AI making headlines for2025 Law firms will need to think more
carefully about AI, as opposed to justbeing a one time purchase as a status
(18:42):
symbol to show we're innovative, toactually think very deeply about their
business model and very deeply abouthow they can integrate AI within their
operations as well as the practiceand delivery of legal services.
From the perspective of, Computers faxmachines or printers being introduced.
People would buy theminitially to demonstrate that
they're ahead of the market.
But at some point, the entire firm hadto get onto a journey, a transformation
(19:06):
journey of adopting PCs laptops computershardware and document management
systems to be effective as a law firm.
It was embedded into the practice.
We're going to see asimilar thing with AI.
except that law firms are in the bestposition to make the most use of AI . AI
is fundamentally driven by what datayou have access to or what ingredients
(19:28):
you can access for the personal chef.
Law firms already have the kitchen.
From the cloud enterprise provider.
They've got access to the ingredients.
They've got access to decades of matter,precedents documentation knowledge
content, all available within the firm.
The key ingredient they'remissing is the talent piece.
If law firms are able to figure out, Howto bring in AI talent into that ecosystem
(19:51):
of integrating technology and AI intotheir practice And into legal services.
Things will be very exciting we'll seea shift from the knowledge management
perspective where it becomes less aboutcurating data to benefit humans so
that humans can leverage it to curatingknowledge and expertise into data so
that machines can also leverage it.
(20:14):
We like to call this intelligenceengineering, where it will go from just
managing and taking knowledge of theheads of people to actually curating
intelligence that can feed machines,which can then subsequently feed humans.
With the trend of agentic AI.
What we're going to see is the riseof an agent agnostic law firm or agent
agnostic enterprises and organizations.
(20:36):
Typically when we speak about agency,we've only really considered humans
to have agency or to be able tomake decisions and complete tasks.
Now with agents, machines are also goingto have some agency to have autonomy,
to make decisions and complete tasks.
What we're going to see potentially inthe future is organizations and all firms
(20:56):
where work comes in and is orchestrated.
Part of that work is offloadedto a human and part of that
work is offloaded to a machine.
It becomes almost as thoughthat work is agent agnostic.
Whether a machine completes that workor whether a human completes that
work doesn't make a difference aslong as the work product gets done.
(21:16):
So from a future, Looking perspective.
It's really important for law firms to goback to those central questions around.
How do you make sure your datais in a good space internally?
How do you make sure you havethe right infrastructure in
place to enable transformation?
And how do you make sure that from astrategic perspective, AI is part of
the charter of the firm's growth anddevelopment, as opposed to just being.
(21:41):
Something which demonstrates movementon the innovation theater stage.
It's just emotion without reallythinking about it in a very deep way.
This is Ari Kaplan speaking with uesIal, the founder of Plco, a UK based
AI consulting firm, helping law firmsand legal teams with AI adoption.
(22:01):
Oise, wonderful conversation.
Thank you so very much.
Thanks to have Mary.
Thank you for listening to theReinventing Professionals podcast.
Visit ReinventingProfessionals.
com or AriKaplanAdvisors.
com to learn more.
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
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.