September 3, 2025 • 14 mins
I spoke with Sagi Eliyahu, the CEO and co-founder of Tonkean, a process orchestration platform, and the maker of LegalWorks, which helps legal teams automate complex processes and time-consuming workflows. We discussed the best practices for legal teams looking to adopt AI, the difference between how legal teams are using generative AI and agentic AI, and misconceptions about the capabilities of AI.
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 speaking todaywith SGI Eliyahu, the CEO and Co-founder
(00:25):
of Tonkin, a process orchestrationplatform, and the maker of legal works,
which helps legal teams automate complexprocesses and time consuming workflows.
Hi sgi.
How are you?
Hey, Ari.
Good to meet you, man.
It's great to see you.
Tell us.
About your background andthe genesis of Tonkin.
(00:46):
So it's been almost 10 years now Tonkin.
I've been doing enterprise software all mycareer and software since I was a child.
I'm a tech guy.
I like to solve problems with technologyand in my experience prior to Tonkin, I
worked in public companies , and executiveroles and I felt a very specific.
(01:08):
Gap , and a conceptthat stay with us today.
Tonkin and kinda like the inceptionof Tonkin, which is business processes
are about people, not about data.
But 100% of enterprisesoftware is about data.
It's always about the ticket.
And if we focus on legal for a second,it is about the matter and it's
about the document and the contract.
(01:29):
But the process is not about that.
The process is about the people thatare involved in it, people that needs
the help, the service, the stakeholders,the internal folks, external folks.
The processes around the people, buttechnology wasn't optimized for that.
So we started talking and early oncalled it orchestration where no
one knew what I want 10 years ago,but now more people kind of hear
(01:52):
that word and get more open to it.
But it's been quite a ride , andquite a privilege to help a
lot of large enterprises andcorporate legal within them.
Are the organizations that youserve concentrated in specific
practice areas or locations?
Not really.
We are a general platform, and we havecertain solutions across an organization.
(02:15):
Legal and finance are the biggest ones.
And so we've been working with Legalsfor at least five, six years now.
And large enterprises on the globalteams, ones that has corporate
legal teams with hundreds ofattorneys are our bread and butter.
How are legal professionals usingTonkin's Legal Works technology today?
(02:39):
So a lot of the.
Gap that we sort of identify early on,which is true generally and is true even
more for legal in the corporate is wehave our processes and our tooling and
the way we like to work and the, the,the policies and guidelines but then we
service the rest of the organization andthey're not familiar with those processes.
(03:02):
That's not necessarily their priority,you can put all the CLMs and all
the processes around the contract,but then a salesperson would still
email you in the middle of the night.
I need this to review theSAP, you know, for tomorrow
so there's essentially a gap betweenthe internal way we go about doing
things and how other folks consume it.
And that's true also foroutside console and others.
(03:24):
So what Tonkin reallyprovides is a way to build.
Processes that are optimized , andpersonalized for those different
stakeholders without the need of it orengineering , or a custom development.
And in a fraction of the cost ofwhat it will take for you to buy
direct solution for each of those.
So it can, be all the way from, specificprocesses like automating a mail room.
(03:49):
Physical mail that comes into the companythrough conflict waivers or onboarding
of outside counsel to general intake ofany legal requests that comes from the
company, from litigations to contractsto HR related to procurement related
so all of the interfacing processes fromthe legal department to the rest of the
(04:10):
organization , and the outside counsel.
That processes is where token sits andallow to optimize the existing tool sets
with the desire and needs of the employeesor stakeholders on the other side.
What are some best practices for legalteams who are looking to incorporate AI
(04:32):
into this process that we're discussing?
The biggest.
Value add that AI brings.
, And something that is not talkedenough about is the type of
problems that are hard to create.
Hard to generate, tobuild, but easy to review.
(04:53):
So if you think about image generation,if you asked me to, draw you in
a certain style I would need tohave skills to know how to draw.
And to do a good job about it.
But at the same time, that sameperson , doesn't even do have any
skills in order to review the pictureand say, this looks nothing like Ari.
And so there's a lot of things likethat when it comes to legal and
(05:16):
maybe , the most obvious one is initialreview of a contract, where to do a
good job about, actually comparing, the document with guidelines and
understanding all the different, thingsthat needs to be reviewed deeper by
legal, by attorney versus somethingwe know that we just can't accept.
(05:37):
Upfront , training, salespeople,marketing or just people across the
company that are not legal professional.
That's a high toll, but AI can doan amazing job about that and just
highlighting this is a bad clause.
We cannot accept it, don'twaste legal time about it.
Versus, those are , fine, but,we can use different language.
So you can really do a initial screening.
(05:58):
That's one side of it.
Then to your question ofwhat is the best practices?
, You need to have that in a motionthat align with how people work to
my previous concept of talking andthe processes are about people.
It's not enough to justhave that technology.
You need that to be aligned witha natural way of how people work.
(06:19):
So if salespeople, send you anemail about can someone review this?
SAPI would want something automatic toalready listening to that email inbox
and then reply to that email inboxwith an initial review by an AI that
says, okay, here's an initial review.
It looks generally fine, so Ipush it into the legal queue and
(06:40):
someone will look at it tomorrowversus here's an initial review.
There are two red alertshere , here's an immediate thing.
You can come back to your partneror, client or customer, to say, Hey,
those are a few things we have to.
Push back on I will wait.
I will send it to my legal full review,but here are a few things we will
(07:02):
already need to discuss and that'ssome examples that we have customers
already, embracing ai just in thepast year for those type of use cases,
they're seeing a lot of, great results.
Given the release of your recentagentic orchestration solution,
what's the difference in how legalteams are leveraging generative AI
(07:23):
versus now increasingly agentic ai?
There's a lot of noise in the market. There's a lot of hype and, and it's
interesting for me to start by askingwhy is there hype about the concept
of agents specifically ? Generative AIbased on the concept of large language
(07:43):
models, l lms, like JG PT and so on isvery powerful when it comes to language.
So you can analyze large text,you can generate text, you
can generate images and so on.
Where agents defer is that they havethe potential to actually collaborate
and be part of the team because they'rebuilt on the concept of you give the
(08:04):
LLM, you give the AI a goal and you giveit tools, and then the AI can go and
actually try to execute the task at handwith the tools provided versus act as
a chat bot in which, it's just kind oflike question, answer, question, answer.
The difference for legal is
(08:27):
can be quite drastic when you compareit to generative AI specifically.
General AI is just a function ofusing l LMS to generate content while
agent can be specialized in certainareas and, and take away tasks.
I'll give you small examples, like,verifying a client I in address with
(08:49):
authorities, that has nothing to dowith generative AI in the sense of
generating content, but you can build anagent that will go and search, relevant,
databases and verify that for you.
If you combine it with an orchestrationplatform or automation platform
like Tonkin, then you can havethat happen automatically as well
.A generative.
AI tool can generate an NDA for you.
(09:09):
An agent that is an NDA agent can researchexisting information about existing NDAs
with that client or third party lookfor information online or in certain
databases about that third party, and thengen and by generating an NDA, but also
(09:30):
maybe sending that NDA for you and so on.
It's really can take a whole set of stepsand jobs and automated, but all along
being something that you can communicatewith , and give feedback to, unlike
automation where you might have a buttonthat can do everything I just said this
time, the agent has those tools, thosecapabilities, but you can interact with
(09:50):
it and be like, oh no, that's not exact.
So you can just talk toit and it'll do it again.
Where do you think legal teamshave the most misconceptions
about AI and its capabilities?
Legal is very heavy knowledge and content.
The first wave of AI with LLMs createdthis notion that, legal is for some
(10:13):
reason at risk with AI coming in.
For me, this is ridiculous.
It's like saying that accountant,people in finance would feel like
the introduction of Excel meantthat now, their job is at risk.
In reality, it just made their jobsignificantly more important because
(10:37):
things that would take them a longtime to calculate by hand , they're
now responsible for the formula andnot, for the single cell calculation.
Legal would feel.
Over time the outcome would besimilar in where those tools would
allow them to strategize around away to protect the business and to
find better ways to create, leveragein the business and center areas.
(11:01):
Not to mention when it comesto, court work and litigation.
The amount of information you need todigest and prepare for a case you'll have
a lot more tools to do that effectively.
But just like any preparation,it's the questions that matter.
What are you asking?
What are you coming up with?
Overall it'll create a lot of positive,and I think that's a big misconception.
(11:25):
The second thing is.
More general, not only for legal.
Most people think about AI still ascomparing it to what they already
know which, is to be expected.
For example, they might think aboutasking something GGPT as the way , they
ask something into Google search.
(11:46):
But the result they'regetting is one answer.
So they assume that must be the rightanswer because I just got one answer.
In Google, you get 10.
So by definition, you kind of getmultiple results for you to make sense
of it and the right way to work , withchat bots like JGPT, for example, is
to do back and forth and verify it.
Second expecting it to work like any othersoftware in which if you do a certain set
(12:09):
of steps, the outcome would be the sameevery time you do the steps, with an AI
agent, if you say, create an NDA withapple, it'll go and create an DA with
Apple, but then you ask it again, createan NDA with Apple, most likely it'll
tell we already have an NDA with apple.
Why do you want me to create it again?
From a mental perspective, we workmore like a person versus a software
(12:31):
that, do exactly what you say.
It would sometimes pushback or ask more questions.
So the expectations fromit is gonna be different.
But the potential of it will become moreclear for people , and that's exciting.
Like any other software innovation.
It starts by being the expert classthat is kind of clicking around,
(12:57):
understanding where to leverage it.
For anyone that lived beforeGoogle search, it took.
Older folks, like half a decade,if not a decade, to, get to the
concept of every question I have.
I'm just gonna go to Google search.
I don't know about your experience,but my grandma is picking up
GGPT , and talking to it already.
(13:18):
It's more similar to the iPhonein many ways where toddlers knows
how to use an iPad touch , andso on instinctually . Naturally.
Ai, the language side of it is the same.
It's actually a lot more natural.
, To chat or talk to a machine, thento fill forms or move the mouse
(13:39):
around or navigate different tabs.
So I think the expectations aroundinteracting with software is gonna
change faster than people realize.
The way that AI agent will evolveis that they will be everywhere
and everything, like all of thesoftware you are used to work with.
It's gonna transform intoa version of an agent.
(14:03):
It'll be more of a question ofwhat would remain not an agent
than, what will turn into an agent.
This is Ari Kaplan speaking with SGIEliyahu, the CEO and Co-founder of
Tonkin, a process orchestration platformand the maker of legal works, which
helps legal teams automate complexprocesses and time consuming workflows.
(14:27):
Sgi, thanks so very much.
My pleasure.
Thank you for hosting me.
Thank you for listening to theReinventing Professionals Podcast.
Visit reinventing professionals.com orari kaplan advisors.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 speaking todaywith SGI Eliyahu, the CEO and Co-founder
(00:25):
of Tonkin, a process orchestrationplatform, and the maker of legal works,
which helps legal teams automate complexprocesses and time consuming workflows.
Hi sgi.
How are you?
Hey, Ari.
Good to meet you, man.
It's great to see you.
Tell us.
About your background andthe genesis of Tonkin.
(00:46):
So it's been almost 10 years now Tonkin.
I've been doing enterprise software all mycareer and software since I was a child.
I'm a tech guy.
I like to solve problems with technologyand in my experience prior to Tonkin, I
worked in public companies , and executiveroles and I felt a very specific.
(01:08):
Gap , and a conceptthat stay with us today.
Tonkin and kinda like the inceptionof Tonkin, which is business processes
are about people, not about data.
But 100% of enterprisesoftware is about data.
It's always about the ticket.
And if we focus on legal for a second,it is about the matter and it's
about the document and the contract.
(01:29):
But the process is not about that.
The process is about the people thatare involved in it, people that needs
the help, the service, the stakeholders,the internal folks, external folks.
The processes around the people, buttechnology wasn't optimized for that.
So we started talking and early oncalled it orchestration where no
one knew what I want 10 years ago,but now more people kind of hear
(01:52):
that word and get more open to it.
But it's been quite a ride , andquite a privilege to help a
lot of large enterprises andcorporate legal within them.
Are the organizations that youserve concentrated in specific
practice areas or locations?
Not really.
We are a general platform, and we havecertain solutions across an organization.
(02:15):
Legal and finance are the biggest ones.
And so we've been working with Legalsfor at least five, six years now.
And large enterprises on the globalteams, ones that has corporate
legal teams with hundreds ofattorneys are our bread and butter.
How are legal professionals usingTonkin's Legal Works technology today?
(02:39):
So a lot of the.
Gap that we sort of identify early on,which is true generally and is true even
more for legal in the corporate is wehave our processes and our tooling and
the way we like to work and the, the,the policies and guidelines but then we
service the rest of the organization andthey're not familiar with those processes.
(03:02):
That's not necessarily their priority,you can put all the CLMs and all
the processes around the contract,but then a salesperson would still
email you in the middle of the night.
I need this to review theSAP, you know, for tomorrow
so there's essentially a gap betweenthe internal way we go about doing
things and how other folks consume it.
And that's true also foroutside console and others.
(03:24):
So what Tonkin reallyprovides is a way to build.
Processes that are optimized , andpersonalized for those different
stakeholders without the need of it orengineering , or a custom development.
And in a fraction of the cost ofwhat it will take for you to buy
direct solution for each of those.
So it can, be all the way from, specificprocesses like automating a mail room.
(03:49):
Physical mail that comes into the companythrough conflict waivers or onboarding
of outside counsel to general intake ofany legal requests that comes from the
company, from litigations to contractsto HR related to procurement related
so all of the interfacing processes fromthe legal department to the rest of the
(04:10):
organization , and the outside counsel.
That processes is where token sits andallow to optimize the existing tool sets
with the desire and needs of the employeesor stakeholders on the other side.
What are some best practices for legalteams who are looking to incorporate AI
(04:32):
into this process that we're discussing?
The biggest.
Value add that AI brings.
, And something that is not talkedenough about is the type of
problems that are hard to create.
Hard to generate, tobuild, but easy to review.
(04:53):
So if you think about image generation,if you asked me to, draw you in
a certain style I would need tohave skills to know how to draw.
And to do a good job about it.
But at the same time, that sameperson , doesn't even do have any
skills in order to review the pictureand say, this looks nothing like Ari.
And so there's a lot of things likethat when it comes to legal and
(05:16):
maybe , the most obvious one is initialreview of a contract, where to do a
good job about, actually comparing, the document with guidelines and
understanding all the different, thingsthat needs to be reviewed deeper by
legal, by attorney versus somethingwe know that we just can't accept.
(05:37):
Upfront , training, salespeople,marketing or just people across the
company that are not legal professional.
That's a high toll, but AI can doan amazing job about that and just
highlighting this is a bad clause.
We cannot accept it, don'twaste legal time about it.
Versus, those are , fine, but,we can use different language.
So you can really do a initial screening.
(05:58):
That's one side of it.
Then to your question ofwhat is the best practices?
, You need to have that in a motionthat align with how people work to
my previous concept of talking andthe processes are about people.
It's not enough to justhave that technology.
You need that to be aligned witha natural way of how people work.
(06:19):
So if salespeople, send you anemail about can someone review this?
SAPI would want something automatic toalready listening to that email inbox
and then reply to that email inboxwith an initial review by an AI that
says, okay, here's an initial review.
It looks generally fine, so Ipush it into the legal queue and
(06:40):
someone will look at it tomorrowversus here's an initial review.
There are two red alertshere , here's an immediate thing.
You can come back to your partneror, client or customer, to say, Hey,
those are a few things we have to.
Push back on I will wait.
I will send it to my legal full review,but here are a few things we will
(07:02):
already need to discuss and that'ssome examples that we have customers
already, embracing ai just in thepast year for those type of use cases,
they're seeing a lot of, great results.
Given the release of your recentagentic orchestration solution,
what's the difference in how legalteams are leveraging generative AI
(07:23):
versus now increasingly agentic ai?
There's a lot of noise in the market. There's a lot of hype and, and it's
interesting for me to start by askingwhy is there hype about the concept
of agents specifically ? Generative AIbased on the concept of large language
(07:43):
models, l lms, like JG PT and so on isvery powerful when it comes to language.
So you can analyze large text,you can generate text, you
can generate images and so on.
Where agents defer is that they havethe potential to actually collaborate
and be part of the team because they'rebuilt on the concept of you give the
(08:04):
LLM, you give the AI a goal and you giveit tools, and then the AI can go and
actually try to execute the task at handwith the tools provided versus act as
a chat bot in which, it's just kind oflike question, answer, question, answer.
The difference for legal is
(08:27):
can be quite drastic when you compareit to generative AI specifically.
General AI is just a function ofusing l LMS to generate content while
agent can be specialized in certainareas and, and take away tasks.
I'll give you small examples, like,verifying a client I in address with
(08:49):
authorities, that has nothing to dowith generative AI in the sense of
generating content, but you can build anagent that will go and search, relevant,
databases and verify that for you.
If you combine it with an orchestrationplatform or automation platform
like Tonkin, then you can havethat happen automatically as well
.A generative.
AI tool can generate an NDA for you.
(09:09):
An agent that is an NDA agent can researchexisting information about existing NDAs
with that client or third party lookfor information online or in certain
databases about that third party, and thengen and by generating an NDA, but also
(09:30):
maybe sending that NDA for you and so on.
It's really can take a whole set of stepsand jobs and automated, but all along
being something that you can communicatewith , and give feedback to, unlike
automation where you might have a buttonthat can do everything I just said this
time, the agent has those tools, thosecapabilities, but you can interact with
(09:50):
it and be like, oh no, that's not exact.
So you can just talk toit and it'll do it again.
Where do you think legal teamshave the most misconceptions
about AI and its capabilities?
Legal is very heavy knowledge and content.
The first wave of AI with LLMs createdthis notion that, legal is for some
(10:13):
reason at risk with AI coming in.
For me, this is ridiculous.
It's like saying that accountant,people in finance would feel like
the introduction of Excel meantthat now, their job is at risk.
In reality, it just made their jobsignificantly more important because
(10:37):
things that would take them a longtime to calculate by hand , they're
now responsible for the formula andnot, for the single cell calculation.
Legal would feel.
Over time the outcome would besimilar in where those tools would
allow them to strategize around away to protect the business and to
find better ways to create, leveragein the business and center areas.
(11:01):
Not to mention when it comesto, court work and litigation.
The amount of information you need todigest and prepare for a case you'll have
a lot more tools to do that effectively.
But just like any preparation,it's the questions that matter.
What are you asking?
What are you coming up with?
Overall it'll create a lot of positive,and I think that's a big misconception.
(11:25):
The second thing is.
More general, not only for legal.
Most people think about AI still ascomparing it to what they already
know which, is to be expected.
For example, they might think aboutasking something GGPT as the way , they
ask something into Google search.
(11:46):
But the result they'regetting is one answer.
So they assume that must be the rightanswer because I just got one answer.
In Google, you get 10.
So by definition, you kind of getmultiple results for you to make sense
of it and the right way to work , withchat bots like JGPT, for example, is
to do back and forth and verify it.
Second expecting it to work like any othersoftware in which if you do a certain set
(12:09):
of steps, the outcome would be the sameevery time you do the steps, with an AI
agent, if you say, create an NDA withapple, it'll go and create an DA with
Apple, but then you ask it again, createan NDA with Apple, most likely it'll
tell we already have an NDA with apple.
Why do you want me to create it again?
From a mental perspective, we workmore like a person versus a software
(12:31):
that, do exactly what you say.
It would sometimes pushback or ask more questions.
So the expectations fromit is gonna be different.
But the potential of it will become moreclear for people , and that's exciting.
Like any other software innovation.
It starts by being the expert classthat is kind of clicking around,
(12:57):
understanding where to leverage it.
For anyone that lived beforeGoogle search, it took.
Older folks, like half a decade,if not a decade, to, get to the
concept of every question I have.
I'm just gonna go to Google search.
I don't know about your experience,but my grandma is picking up
GGPT , and talking to it already.
(13:18):
It's more similar to the iPhonein many ways where toddlers knows
how to use an iPad touch , andso on instinctually . Naturally.
Ai, the language side of it is the same.
It's actually a lot more natural.
, To chat or talk to a machine, thento fill forms or move the mouse
(13:39):
around or navigate different tabs.
So I think the expectations aroundinteracting with software is gonna
change faster than people realize.
The way that AI agent will evolveis that they will be everywhere
and everything, like all of thesoftware you are used to work with.
It's gonna transform intoa version of an agent.
(14:03):
It'll be more of a question ofwhat would remain not an agent
than, what will turn into an agent.
This is Ari Kaplan speaking with SGIEliyahu, the CEO and Co-founder of
Tonkin, a process orchestration platformand the maker of legal works, which
helps legal teams automate complexprocesses and time consuming workflows.
(14:27):
Sgi, thanks so very much.
My pleasure.
Thank you for hosting me.
Thank you for listening to theReinventing Professionals Podcast.
Visit reinventing professionals.com orari kaplan advisors.com to learn more.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
Cardiac Cowboys
The heart was always off-limits to surgeons. Cutting into it spelled instant death for the patient. That is, until a ragtag group of doctors scattered across the Midwest and Texas decided to throw out the rule book. Working in makeshift laboratories and home garages, using medical devices made from scavenged machine parts and beer tubes, these men and women invented the field of open heart surgery. Odds are, someone you know is alive because of them. So why has history left them behind? Presented by Chris Pine, CARDIAC COWBOYS tells the gripping true story behind the birth of heart surgery, and the young, Greatest Generation doctors who made it happen. For years, they competed and feuded, racing to be the first, the best, and the most prolific. Some appeared on the cover of Time Magazine, operated on kings and advised presidents. Others ended up disgraced, penniless, and convicted of felonies. Together, they ignited a revolution in medicine, and changed the world.