What if the most powerful AI in your practice still isn't the thing winning your cases? Sara Cordrey, CEO and co-founder of the legal AI platform Parambil, built with chief medical officer Dr. Ralph Horwitz, argues that technology is leverage—not a substitute for a lawyer's judgment or for the medical record itself. Drawing on her McKinsey work analyzing gaps in patient care, she explains how purpose-built AI helps trial lawyers rip through records, reorder how cases get worked up, and surface insights sooner—while never leaving the primary source behind. Host Ben Gideon digs into build-versus-buy, data security, and where legal AI is heading next.

Learn More and Connect

☑️ Sara Cordrey | LinkedIn

☑️ Parambil | LinkedIn | YouTube | Instagram

☑️ Ben Gideon | LinkedIn | Facebook | Instagram

☑️ Gideon Asen on LinkedIn | Facebook | YouTube | Instagram

☑️ Rahul Ravipudi | LinkedIn | Instagram

☑️ Panish Shea Ravipudi LLP on LinkedIn | Facebook | YouTube | Instagram

☑️ Subscribe: Apple Podcasts | Spotify

Produced and Powered by LawPods, a podcast marketing agency for law firms

Sponsored by SmartAdvocate, Hype Legal, Expert Institute, and Steno.

Welcome to Elawvate,

the podcast where trial lawyers Ben

Gideon and Rahul Ravipudi talk about the

real issues that come with

the fight for justice.

So let's find inspiration in the

wins. Let's learn from the losses.

But most of all,

let's keep learning and getting better

and keep getting back in the ring.

Are you ready to elevate your own

trial practice, law firm, and life?

Let's get started. Produced

and Powered by LawPods.

Hey, it's Ben.

Raul and I started this podcast because

we love hanging out with fellow trial

lawyers and sharing ideas

that can make us all better.

And both of our firms also regularly

collaborate with other lawyers across the

countries in cases where we can add value.

If you're interested in collaboration or

even if you just have a case or an idea

that you want to bounce

off us or brainstorm,

Raul and I are going to be hosting

confidential case workshops the first

Wednesday of each month.

So here's how it works.

If you have a case or an idea that you

want to talk about or brainstorm with us,

just send me an email to [email protected],

E-L-A-W-V-A-T-E.net,

or go online to elawvate.net and

submit a case workshop request.

We will schedule you for a confidential

30-minute Zoom meeting where we can talk

about your case to see if we can help.

If you feel like there would be good

value in collaborating on the case

further, we can talk about that. If not,

that's okay too.

We enjoy helping other trial

lawyers because we know

someday you'd be willing to

do the same for us if we

needed your help. So again,

if you're interested in

workshopping your case with us,

just send an email to [email protected]

or fill out a caseworkshop request at

elawvate.net. And Rahul and I will

look forward to chatting with you soon.

Today's episode of the Elawvate Podcast

is brought to you by Steno. Rahul,

you guys work with Steno.

Steno is the best in

court reporting services,

not just in court reporting services,

but even some of their technology tools.

We're talking about AI a little bit on

this podcast and their transcript genius

where they can summarize and take

interrogatories based on deposition

transcripts is so useful. If you

haven't tried it, definitely try it.

Now we're brought to you by HypeLegal.

HypeLegal does digital marketing,

web development for trial firms.

It's owned by our good friends,

Micah and Tyler. They recently

redeveloped our firm's website,

so you can check our

website out. If you like it,

give them a call and they can

help you out too. And finally,

we're brought to you by Expert Institute.

Rahul, you guys work with them, right?

We both use Expert Institute because you

always need to be cutting edge in the

experts that we use in our cases.

Going to the repeat experts every single

time is going to make you a lesser

lawyer and you always want to keep up and

the best way to do that is with Expert

Institute.

Welcome to the Elawvate

Podcast. I'm Ben Gideon.

Rahul is off today prepping for trial.

So I'm going to do this episode solo,

but really not solo at all because I'm

really thrilled to have Sarah Cordrey

join us today. Sarah, as I understand

it, is the founder, the CEO,

and the moving force behind a

new AI product that I've grown

to rely on and use

extensively in our practice.

And I can personally vouch for it being

a real game changer for trial lawyers

and particularly those of us who

do medical malpractice cases,

I'm sure other types of complex

litigation. So first of all, Sarah,

welcome and thank you for the product

you've developed. It's really,

really been of enormous

value to us in our practice.

Ben, thank you for having me.

Excited to be on and honored that great

trial lawyers like you are finding value

out of an engineer's product.

Yeah, and I don't say

that lightly. I mean,

we've used a lot of different products,

including some I've panned on this

very show. We have very high standards,

but the platform you've

created, it's called.

One thing that's tricky

is how to say the name.

So hopefully you can

clarify this for good.

It's spelled P-A-R-A-M-B-I-L.

And in our office,

some people call it Parambil and some

people call it Parambil. But what is it?

So we know for sure. The.

Crazy thing is it's

either. I call it Parambil,

but I think the days of companies

trying to name everything themselves are

over. We let our clients

name their AI agents.

Some call our platform Par, some

call it Bill, some call it Fabio.

So I think choose the name that

works for you at the end of the day.

Damn it.

I was hoping you were going to prove

me right and someone else in our office

wrong about that. Where did the

name come from? I'm curious.

So Dr. Ralph Horwitz, my co-founder

and chief medical officer,

was very good friends with the

author of an Oprah's book club book,

The Covenant of Water.

It's by Abraham Verkees and is

a book that's about this long.

It's the size of a Bible

for those who are listening.

And I was reading it during the early

parts of building this company with my

co-founders, and the setting

of that book is Parambil.

And it's a place in India and

it's a tiny town when it starts.

But by the end of the story, which

is really a medical story in itself,

it's this bustling hub of energy

and excitement and activity.

And so it felt like a natural name

for a company that went from a

tiny, tiny bit of nothing to something

really powerful and full of energy.

Oh, the name's even better now that I

know the backstory about it. That's cool.

So tell us a little bit about

yourself. What's your background?

How did you get into this industry

and have interest in trial, law,

AI type platform?

I'm originally from Miami, live in

New York now with husband and kids,

so did the opposite of what

everyone else did during COVID.

I stayed in the cold. By background,

I'm an engineer. And interestingly,

I have spent the majority of my career

focused on similar problems that I

solve now. I remember I was

in an Uber. I was maybe 21,

22 years old, and I was working for

McKinsey and my code had just come back.

The results were here. And I thought

I must have messed up the code.

I must have made a mistake. I've been

building these algorithms for weeks, days,

nights, and there's no way that

the results I'm seeing can be true.

And what I was doing was I was working

with this group called Medical Unmet Need

at McKinsey,

and we would take these massive

population level data sets across the US

and analyze care to understand where

patients weren't getting the care they

should. And then we would

give that information to

pharma companies and they in

theory would go send their reps and go

help talk about products and all these

other commercial components.

But the goal was find patients where

their care is deviating from the

guidelines, which probably as a med mal

attorney sounds very familiar to you.

And so when we were doing these analyses,

we would find these huge clusters,

thousands of people who had clear

symptoms that they needed care,

they needed an escalation,

but they weren't actually getting

the care they needed. Again,

despite going to a doctor's office

multiple times. And so as a young,

naive person, I thought, how is

this happening? This can't be true.

And we would talk to

doctors and we'd ask them,

what's going on in these

offices? And at that time,

my belief was they just don't have the

information they need. They don't have

the full patient story. Fast-forward,

we keep working on this,

and Ralph comes to me and says, Sarah,

I think that the old machine learning we

used to use is a thing of the past and

this new AI, this new generative AI

is going to unlock medical records.

And with that,

a really rich data set that will give us

insights on patient care that we could

never have before. And so we started

off, we left McKinsey together,

which was a little crazy

at the time. And we said,

how can we bring this new data set to

better tell a patient's story in the

doctor's office? Obviously, we're

talking today, so that didn't work.

We quickly realized that the doctors

could ask the questions that they were

missing at the time. They could order a

lab result, they could fix all the gaps,

but Ralph had been an expert witness

and he started to talk about how in his

experience when there's harm in the US,

that harm is documented in the medical

record. And when you're in a courtroom,

when you're the other side, you can't

just go request the lab results.

You can't just go ask the questions

and fill in the gap. Those documents,

that data become the basis for your

story for the facts of the case.

And so that was three years ago now,

and we've been deep in the legal space,

learning and building ever since.

How did you make that transition to

decide to go into the legal space?

And what was the first step of that? It.

Was a lot of phone calls and a lot of

LinkedIn. Ralph came to me and said,

"I remember when I was an expert witness,

Ralph was head of

medicine at Yale Stanford.

He was involved in big mass tort

cases, med mal cases, you name it.

He did it early on in his career,

and he was recounting these bankers

boxes full of medical records,

full of literature from the

epidemiology side. And he said,

I'm pretty sure this is still done

by hand." And of course, as a young,

stubborn person, I said, "Well,

let me go find out if that's still

true." So I personally called hundreds of

personal injury firms across the US.

I cold called them and tried to get

them to tell me a little bit about their

problems and their pain

points. And from there,

started to really get a

sense of the challenges that

these firms were facing. So

it was a lot of listening in the early

days, and it still is a lot of listening.

Yeah. I mean, you can tell

that from using the platform,

a lot of thought has gone into the

way in which we analyze and approach

cases because the AI seems to

mirror the way we think about

a case. Can you just kind of give us a

primer for those who haven't used this

platform or understand how

it works and what it does?

Can you maybe give us the 10,000 foot

overview first before we drill down into

some of the details?

Definitely.

And feel free to jump in as I explain

it because you probably know it better

than me at this point, but the kind of

basic tool is a case intelligence tool.

And the idea is a case comes in to Ben's

office and we're deciding whether or

not to take the case.

So the team dumps all the data they

have about the case into the system,

whether that's medical

records, intake transcripts,

anything they have that's documents.

And the system helps build an initial

sense of is there a case or not?

It looks at standard of

care, identifies deviations,

different analyses depending on the type

of law that that case is being reviewed

for. From there, if

someone takes on the case,

it jumps into the pre-lit mode and you

start really working up and digging into

the case. And their firms are able to

build the documents that they need,

complaints. We've even heard of HIPAA

authorizations being drafted in there,

and do the work to get that case moving.

And that case is continually updated,

refreshed as new data comes in,

and then moved all the way to the

settlement or outcome phase where they're

closed out.

So really it's a tool that

allows firms to operate to a

higher level,

and it arms them with the intelligence

they need to build case strategy and

ultimately win in court.

And just to put people's minds at ease,

what's the kind of security situation?

Is the platform secure,

HIPAA compliant, so forth?

The platform is HIPAA compliant.

It is also SOC2 compliant.

Those certifications

are thrown around a lot.

What that means is we have an

independent audit firm come in,

review all of our

processes in our systems.

They actually bring in an

outside cybersecurity company

that simulates a tax on

our platform. And every

year on a consistent basis,

we make sure that we're as

secure as is humanly possible.

I think that's really important.

But I think the other thing that matters

from a security perspective is what's

your day-to-day?

And we use AI agents to actually

simulate a tax on our system

weekly and every time we make an

update to the code base so that we're

constantly preparing for how do we

fortify our client's data as much as

possible. That's a great question.

Can you tell us a little bit more about

the company? You mentioned starting out,

you and your colleague leaving McKinsey,

which I'm sure was a leap because I

imagine that was a pretty secure and

comfortable job. So you're taking a

leap of faith there to start something.

How has that gone and where are you

in your company building journey now?

Yeah, I feel a kinship to our

clients because we work with many

entrepreneurs, people who've

started their own business,

who've leapt into the world and

realized it's on you to create value.

It's on you to make this business

important and make this worth something.

And so I think our business has grown

alongside many other businesses,

which is special. We've been at

this for over three years now.

We've raised $8 million

in funding from top VCs.

We have a massive network of top

litigation firms on our platform in the

many hundreds at this point.

And you asked about the change.

When we started, we were doing

a lot with medical records.

That was our bread and butter,

and it still is, right?

That's the heart of the case.

That's never not the focus,

but the platform has expanded a

lot within the past three years.

And now it's doing everything

from drafting documents

to what we'll be launching

by the time this podcast comes out,

which is those more CRM-like capabilities

where you're actually automatically

pulling data, you're giving admins,

firm-level views. So it's really gone

from a little piece of the puzzle,

a medical expert within the case,

to a full engine that is making

every law firm we work with an AI

native law firm.

I'm really interested in your thoughts

about where all of this is heading.

I mean, it's evolved so

fast. Just a year ago,

we couldn't have had this same

conversation. And even your platform,

which we've now been on for, I

don't know, four, five months,

and using it extensively has

improved dramatically over that time,

and I imagine we'll

continue to do so. I mean,

firms are accustomed to having

different systems such as you

alluded to a CRM. We

use the Microsoft Suite,

so we're using all of the

standard products there and

SharePoint to store some

and use some documents along with

our CRM to do different things.

And then we have Parambil to do more

of the detailed case analysis and

litigation workflow there. But where

do you think all of this is heading?

I mean, it does seem it would be more

efficient to have one platform that can do

everything.

At the end of the day, our job is to

create value for customers, right?

To help you win, to help you do

things in a more efficient way.

And you hit on something big,

which is there's all these

different places where data lives,

where information is stored,

where work gets done.

And more and more the

ask from our clients is,

can you pull all of that into one

central system so that our firm's living

documents, our data,

all this information becomes a

rich repository that we can build a

generational law firm from?

So I think where every player is racing

to right now is how do we become the

operating system for a firm?

And we do that by making

every hard problem that you

face today easier to solve

within Paramble's ecosystem.

Obviously,

there's platforms like Harvey that have

become big within the corporate law

setting, big law,

and then there's stories about some law

firms that have exited that and they're

investing money to build

their own internal systems,

maybe built on other

platforms like Palantir.

So there's kind of the go join in

with others with a platform that

is being developed by an outside

vendor versus the building something

internally that can be unique to the

firm itself and maybe give them their own

sort of potential competitive advantage

if they can do something better than

someone else. But dubious of that since

we're in the business of practicing law,

not data science or computer

engineering or software engineering.

But I don't know, where do

you see that battle heading,

the kind of internal proprietary

systems versus adopting

or your platform becoming kind

of the standard across the board?

I used to be part of the, I

don't want to say problem,

but I used to be part of the problem

when I was at McKinsey where you'd have

consultants come in and

they'd say, Big Pharmaco,

why don't you build this

proprietary software yourself?

Doesn't that sound exciting?

Doesn't that sound like a secret sauce

that you can tell all of your investors

about?

And what it usually resulted in was

a very expensive build followed by an

ongoing maintenance that never ended.

Because to keep up with innovation,

to keep the platform on the bleeding

edge requires a lot of work and a

lot of time. I think it can be

done, right? Kirkman and Ellis,

I think is what you're referring to.

And I think they're

spending 500 million on it,

so you certainly have budget

to do it. But the question is,

are you winning trial because of your

software or because your skills as a trial

lawyer? And this might be a controversial

take because I'm a builder of

software. I think our software helps

these firms get bigger outcomes,

see information that they

might not have seen as quickly,

but you are the ones

that are winning, not us.

And I think it's a bit hubristic to

imply that Parambil's the reasons Ben is

getting that multimillion dollar

verdict. Sure, Parambil's helping,

but you're the one that's

winning at the end of the day.

Take your playbook to the next level by

joining a legal network of all stars at

psrplaybook.com.

Psrplaybook.com is a free digital

legal library created exclusively

for plaintiff lawyers by the nationally

recognized personal injury law firm of

Panish Shea Ravipudi LLP and

features video content showcasing the

tactics, strategies,

and concepts implemented by some of

the industry's most successful trial

attorneys. Psrplaybook.com promotes

learning, shared knowledge, collaboration,

and best practices.

Get in the game and see for yourself

by joining our national network of

plaintiff lawyers.

Subscribe now at psrplaybook.com

and learn from the best.

In addition to Paramble, which is what

I've always called it, across our firm,

we use ChatGPT.

Everybody has an account for that

and they use it for certain things.

I've actually developed a whole

Claude-based project system that I use

extensively for other things.

And sometimes I will run some

of my cases through Claude or

I'll have Parambil develop a

detailed research memo using the

proprietary nature of it,

and then scrub it for client identifiers

and then run it through Claude or

Chat. I'm just curious, I mean,

obviously my sense is that all of these

platforms are using the same large

language data sets. That

seems obvious to me,

but is there still a use case for

using different platforms for different

things? And if so, how do you

break that down for folks?

Yeah, that's a great

question. And you're spot on.

We use whatever foundation models

are the best for the task at hand.

So we work closely with Anthropic's

team, OpenAI's team, Google's team,

and we're constantly evaluating and

benchmarking these models and figuring out

what's the right spot to put

each model within our workflow.

And let me just say, Claude is

amazing. I use it all the time.

We have a lot of respect for that

company and the work that they put out.

I would say what's different is

Claude is a general purpose tool.

We are very tailored to the types of

work that you and other personal injury

attorneys do.

And so that purpose-built component

and forward-looking component helps us

iterate very quickly towards

the work that you need done.

So chances are if it wasn't

in Parambil two weeks ago,

it will be in two weeks.

We're moving quickly on those use cases.

And the way I would think about it too

is let's say you're going into trial.

Do you want a part-time attorney on

the case or do you want someone who is

spending day and night focused on the

same problems and winning that case?

Frankly, if Anthropic were to turn to

this space and try to do the same thing,

they'd probably do an incredible

job. They're just not. We are.

And so we're purpose-built,

we're forward-looking,

and we're building more and more.

So chances are if you're doing it in

Claude today and talk to one of our

engineers, it will be in our

platform soon. How do you think, Ben,

about these Claude workflows

as a business owner?

Does that scale across your

organization or do you run into hiccups?

It's a great question. I

mean, I use Claude personally,

but we have not ruled it out across the

board to the rest of our office yet.

Everybody else is on chat

and we use that more for

routine tasks, administrative

tasks, back office functions,

not for analyzing records

or uploading client data.

It's been great for that.

I mean, for example,

there could be if we're doing

intake calls, as we do many,

many intakes with people who have

potential medical malpractice cases,

we're not going to enroll in a parable

case for somebody who just happens to be

on the phone.

But our intake person can

input data from the call into

chat without any client-specific data,

but just kind of what are the issues?

They had a hip replacement, they had

bleeding, they had complications,

and quickly get guideline standards,

things that can help them

during that phone call,

ask what are the questions I need to ask

during a 10-minute call to drill down

and clarify whether this might be a case

or not. So we use it more like that,

and it's really helpful, I think,

in extending the reach of people who

don't have highly technical backgrounds in

terms of medicine or maybe aren't

as experienced and know what to ask.

Very much accelerates the learning

curve of people doing medical cases,

which is it can take a long time to

become serviceable and being able to

issue spot and analyze

cases. But we haven't rolled

Claude out across the board.

We had been marketing for a chief

AI officer with the idea that we

were going to build out our own kind

of AI structure for workflow and

operations. And we're still

very much interested in that.

We interviewed a really great

candidate and offered her the position,

and then she got a better

offer at a different company.

And so we kind of had to go back to ground

zero on that. That's why I asked you

about the build your own versus

rely on third-party vendors.

I've had somewhat of a change of heart

just because the progress I've seen on

your side has been so great that a lot

of the problems we identified that hadn't

been addressed are now being solved

by other people in rapid fashion.

And obviously you're investing a

lot of time and money to do that.

Why should we duplicate that

when it's not necessary?

What are you hearing from the lawyers

who are using your platform in terms of

what people love about it and what are

the next things people are looking for

they wish it could do

that it can't do today?

Flipping your question order.

By the time this is out,

there are two big things that I think

your firm and others will love. The first,

and clearly I'm not a marketing

person, I'm an engineer.

So if you want to rename these.

By the way, your marketing person,

I think his name is Jordy, right?

The guy I originally dealt with.

I really enjoyed working with him.

He was great, very personable,

not a hard sales guy,

just really nice relational person.

And I've dealt with so many of these

hardcore salespeople from different firms

trying to push products on you. I

really appreciate that. He was great.

I'm glad to hear. As I was walking

out of the office yesterday,

my account executives were discussing

whether ACOG guidelines were a little too

conservative.

Not your.

Typical sales guys. A.

Little water cooler conversation about

the ACOG guidelines. That's nice.

But the two things that I

think you'll find exciting.

So number one is called Power Query. And

it is the idea that you can go in, Ben,

as your firm leader, and ask any

question you want across all cases.

And you can use this to

better steer your ship.

You can also use it to

find potential new cases.

You can scan all your med mal cases and

see if any qualify for a mass tort case.

You can scan all of your data and make

sure that complaints are being completed

in a timely fashion. And frankly,

I can't imagine what else people will do

because we launched these products and

the uses explode. So

that's one, Power Query.

And then two speaks to this admin

component that you were talking about.

So we are integrating with calendars,

with transcription services,

with anything you need,

and we are building workflows

that let you run every part of

your case through with the brain of Ben

Gideon guiding it basically. So you can

go in and you can take a MedMal workflow

and you can tailor it maybe based on

what you've built in Claude and pull

in the skills, pull in the templates,

pull in the documents you want

drafted on demand, pull in the tasks,

and facilitate the way work

gets done for your team.

And also in some ways kind of watch it

all happen and make sure it's happening

to the quality and level you would expect.

That's amazing. Now, what

is that going to be called?

Is there a name for that function?

People are calling it AICRM, but

if you help me come up with a name.

That's a terrible name. It's got to have

something more catchy than that. Yeah.

I'll come up with one. I tend

to be really good at that,

although the name of this podcast

doesn't seem to reflect that,

but that's Raul's. We blame him for that.

Yeah. I mean,

what you just described is exactly what

we are trying to accomplish with the med

mal review section of our Claude workflow,

which is to take the collective

knowledge of people that have decades of

experience doing this

and allow their brains,

the way they think about the

case, to translate to the newer,

younger lawyers and potentially

paralegals so that they're getting the

benefit of our knowledge and be able

to leverage that across the firm.

So that would be amazing to have that

and really to make it personalized

because I mean, we're all, I'm

sure, very arrogant. But I mean,

I've spent a lot of time doing this work

and I would like to think I've learned

quite a bit over the years that has

unique value and that it would be great to

preserve that and be able to hand that

off to others and have them benefit from

it.

So I'm excited to hear about that.

You probably got to that point because

you heard from lawyers out there who

wanted to accomplish that goal, right?

Exactly. Lots of listening.

And I think what's also important is

because we've gone back and forth with

continuously learning system. Our

system doesn't train on any client data,

it doesn't train on any client

interaction. We do not do that.

And we've thought about whether opting

into training would help or hurt a law

firm. And at the end of the day,

you do not want your firm brain

learning from the mistakes or the bad

habits of maybe your newest employees

or the employee that does something a

little bit crazy.

And so part of how we've been designing

this is how do you allow the people with

all the training,

the people guiding the ship to put in

their learnings without propagating bad

habits, as they might say.

Right. I mean,

one of the things as a firm owner

and leader that I constantly worry

about is that in the old days,

we would review paper records and

flag them with Post-it notes and

highlighters.

And then at some point that transitioned

to paperless and we would do the same

thing in PDF form. And we still

do and bookmark our records.

But that process kind of gives you,

it structures your brain to have a

physical understanding of a record

and what the different

components are and what should be

there,

what maybe you can readily identify

as missing that you should expect to

find, how things are organized.

And just like any system,

a record in a way reflects

an organizational structure

and a thought process of

whoever's developing that and the people

that have input into it, the doctors,

nurses, et cetera. It's

more than just paper,

it's a reflection of a system and work

process and thought process. I mean,

even just the selection of what to

go in, what's in and what's not,

or if you have dropdown menus,

what are the options on the menu to

choose from or what populates a field

automatically versus all these and what

data is tracked in a flow sheet and what

data is not tracked in a flow sheet?

And all those things play into your

understanding of how medical care is

delivered in a case.

My worry is that jumping right to

a product like Paramble for newer

lawyers who don't know that,

they don't have the understanding of

what underlies that. But it's also,

I mean,

do you need to train somebody

to do complex division in

their head if they have

a calculator? Is that a good use of

anybody's time? No, it probably isn't.

So just balancing that. I mean,

we can be much more efficient now

and accelerate people's learning,

but where is the balance between

training and real understanding

versus just maybe getting to an answer,

but not understanding how you got there

or why? And does that create extra risk?

And I'm just curious,

you obviously know all of this because

you started out doing the processing of

the medical data in a different

world, kind of like we did,

and then have evolved to this world.

So where do you see the balance there

of real understanding on the human

level and how to interact with the

AI if the AI's understanding is now

exceeding that of some of the

people that are new to this work?

This question takes me back to fifth

grade when you're learning about primary

versus secondary sources.

I think what happens with technologists

is technologists get over eager and they

go and take everybody into

the new world very quickly.

And I remember it was probably

three or four months ago,

we were talking to a great

firm in Connecticut and

they were walking me through

their process. And I felt like I had

been hit over the head with the obvious,

which was as much as AI is

going to help all these pieces,

you're never going to leave

behind the medical record.

You shouldn't leave behind the medical

record. That is your primary source.

That is the fact.

There is nothing like the primary

unabridged version of that data source.

And that actually prompted

us to go back and say,

"We've been spending all this time

thinking about how to ask questions to the

record. What if we just

organize the record in a way

that lets someone go to the

key pieces of the case faster,

but gets them back in that primary

document?" And so to your question,

I think there's a little bit of a

slingshot where you go really far in one

direction, but then you

pull back and you say,

"What are the different pieces that are

needed for this case?" And the medical

record is one that you cannot replace

and you cannot ignore under any

circumstance.

That's one thing I really like about

your platform is they're things like you

can just look at all the vital signs and

we could do that before they're on page

55 and then page 1022.

So you go through an 8,000 page record

and you can flag all of the blood

pressure findings. But with Parable,

you can just say there's a simple flow

sheet that shows you all of the blood

pressure findings at every

point throughout the chart,

and it can even give you a graph to

show those and you can easily spot

deviations or abnormalities.

And you can do that with a lot of the

different lab findings, vital signs.

It just gets you to that answer very

quickly and organizes in a way that's

really useful.

I think you can always click on the

hyperlink and go back to see it to

confirm that that's actually the accurate

reading that's depicted in the record

if you want to confirm it. I've stopped

doing that as often because I mean,

it's always accurate.

And there are things that are sometimes

not accurate that you want to check,

and that tends to be more of a

conclusion or an editorial comment

on something that's in the record.

But generally the actual data,

like the vital sign was

125 over 76 or something,

that's going to be accurate

if it's listed there.

I'm curious about radiology

and fetal monitoring.

So if you have a birth injury case,

you can elect to choose a

birth injury case in Parambil.

So it will do an analysis of all the

electronic fetal heart monitoring for the

baby.

But my understanding of what it's

doing is just reflecting what the chart

indicates the findings were.

So if there was a documented deceleration

or acceleration or category one,

category two, whatever it may be,

it will reflect that. It will give you

the specific timestamp from the record

and then what the findings were

on the fetal monitoring strip.

But what it's not doing is reading the

squiggles on the page and interpreting

that the way a maternal fetal medicine

doctor would. Similarly with radiology,

I think it can tell you what the

findings are based on a report,

but it's not reading the imaging

the way a radiologist would.

And it might be a lot to ask of course,

but it would be extremely useful at

some point for the technology to be able

to read and interpret

x-rays, MRI studies, CTs,

electronic fetal monitoring.

And I do think that technology

is available at some level,

though probably very expensive

at the medical level,

but do you see there coming a point in

the near future where the platform could

do that?

Yeah, great question. So

one, fetal monitoring strips,

and then two on the radiology side

imaging and having a perspective there.

So on the first one,

the birth injury template has evolved

to actually read those squiggly lines.

The challenge with fetal monitoring -.

Well,

that's funny because I asked Parable the

other day because I had it do that and

I'm working on a birth injury

case with it. And I said,

"Are you interpreting the squiggly lines

or are you just telling me what's in

the chart?" And it said, "I'm not

interpreting the squiggly lines.

I'm just telling you what the nurses or

somebody else documented in the chart.".

Sometimes these agents are a

little too scared of legal battles,

but what's basically happening on our

end is there's a separate processing

pipeline for these fetal monitoring

strips that goes through and extracts that

information and then analyzes it based

off of ACOG guidelines to try to get a

sense of what do you think is a true

deceleration versus acceleration.

And then within the notes

section of the platform,

it's actually trying to match up what

was happening from the record with what

was happening in those

fetal monitoring strips.

Because there's plenty of funky

readings on a fetal monitoring strip.

Someone gets up and goes to the

bathroom, someone is readjusted.

And so trying to tie it up to where

was an intervention truly needed?

And so in the notes, that's where

the model's trying to do it.

We have a different agent

than the Ask AI system,

and maybe we need to tell it

that that's what's happening,

but that is what's

happening behind the scenes.

Okay. And if you go to the fetal

monitoring strip tab and look at that,

that's where you're going to find that?

Exactly. It's pulling all the

data out and then analyzing it.

And then you asked about images.

So we recently started processing

images within the platform,

and that created a host of confusion,

a whole world of confusion because

people thought we were handling imaging,

but it really incentivized the

engineers to solve imaging.

And so they're working on DICOM files and

how to process something that is many,

many, many thousands of images in one,

and that should be launched

within the next month or two. Wow.

That would be very useful. I mean,

even just to get a second opinion or

to have a sense for what might be in an

image, because as lawyers,

we're not really skilled as you

can imagine it reading. Now,

sometimes it's obvious, but

a lot of times it isn't.

And frankly,

you can show fetal monitoring strips

to two experts and get two different

responses.

So there's a level of art and science to

this too that we don't claim to solve.

Well, that's great.

What are other things that you guys are

working on now that we might have to

look forward to in the future?

Lots of cool things. Let me

think of what's shareable.

A couple things that you might find

helpful. Co-counsel relationships,

facilitating work when you have multiple

firms together is something that we

just launched. So firms are linking up.

Something in the physical world that

we're doing is a community event in

November that is Paramble customers only.

And the goal is to innovate.

It's to talk to other incredible attorneys

and understand how they're using AI,

what's working,

what are all the things that you could

be doing with AI that you might not even

realize are possible, like HIPAA

authorizations or a bill of particulars,

and then to inform the Parambil roadmap.

So we will have engineers sitting with

our customers as always to listen to them

and understand what they need at a firm

level so that we can make sure that

Parambil is continually evolving as

the technology evolves and as the

needs evolve.

Some of AI is now table stakes,

but we're always pushing on how can

we be better? How can we do more?

How can we create more magical experiences

for the great people we work with?

We do a lot of medical cases obviously,

and we use Parable for

every one of those cases.

It strikes me that it would also have

value for other kinds of catastrophic

injury cases on the damages side for sure,

because you're analyzing the records

and trying to build the damages case to

some extent, medical

records or other things.

Is there a plan to expand

the functionality and

learning out to other types

of plaintiff's cases?

Great question. Yes.

Part of our workflows and our templates

let firms customize the results to what

they need. So we already have

personal injury, motor vehicle cases,

we have trucking templates, we have

mass torts. Any mass tort you can name,

we probably have a template for it.

So the goal is really to

be this flexible system.

I think maybe a mistake on our part is

sometimes we harp too much on the medical

records themselves without focusing

on all that the platform can do.

But that's just because

again, it's where we started.

It's part of the case and we're never

going to stop trying to make that piece

better, but we are servicing

many, many case types now.

And as the case goes forward,

you start to develop the record

with depositions and discovery.

Are you finding people are adding those

documents to the case file and what

are the benefits of doing

that? What can you do with it?

Because we've done less of that.

We use it more to kind of build the

structure of what the legal theories are

in the case. Also,

part of it is it hasn't been around that

long and we haven't used it for that

long. And most of our cases are at

the early phases where we've used it.

So they're all now moving through the

litigation process. What are people doing?

Are you seeing people upload their

discovery documents, depositions,

and then what are the things that can

be done to manipulate or utilize those

through the platform?

The way AI is helping attorneys rip

through data is changing the way attorneys

practice law, even from

a sequence perspective.

We have an attorney down in Florida who

has stopped taking in-person depositions

because she has Parambil up for every

deposition and is asking questions,

going back and forth and creating much

stronger results because she's using the

platform in tandem.

We also have firms that are dumping

in audit logs and they're actually

requesting them earlier now from a

case perspective because now they have

something that can power through them

and find them the insights. Before,

it was a trade-off of,

do I have a paralegal spend two weeks

going through a giant Excel or do I focus

on something else that I need for trial?

And so we see the types of data being

pulled up earlier into a case because they

can result in more. And you asked

about other types of data. I mean,

our average case now has over a thousand

documents because people are dumping in

everything. I had a salesperson run into

my office the other day and say, Sarah,

I need to tell you about the

coolest thing I just did.

And he had a case with a customer

where the person devolved in care and

ultimately needed a hyperbaric

chamber to resolve the injury,

but the hospital's hyperbaric

chamber was down. So the salesperson,

Sal and the attorney were sitting there

and they started dumping in all the

hospital's policies and

procedures, the guidelines.

They're starting to try to figure out

what do you do when a hyperbaric chamber

is down? What is the policy

here? How should you act?

How should you treat this injury?

And by the end of 30 minutes of them

asking questions and staring at it,

they built up a really strong case.

And that's maybe not the most obvious

use case of Paramble, understanding

the hyperbaric chamber outage policies,

but we are seeing things like

this analyzed on a daily basis.

My Facebook feed tells me I can buy a

personal hyperbaric chamber for my home

for only $60,000.

I don't know what it is about me that

makes them think that I'm in the market

for one of those, but maybe I need one.

The challenge with this type of

law, you get a lot of targeted ads.

Yeah. So what you just described

in putting a lot of data in and

an audit trail, for instance,

I had been putting my audit trails

into Claude and having you do that,

but that you're saying that Parambil

would process audit trails and do the

analysis on that's been

educated on how to do that.

We see them with trucking cases

too, right? Trucking cases,

you have all of these activity logs for

messages and interactions that in some

ways look like an audit log and

are just as hard to run through,

but the AI can help not just

review it, but to your point,

understand how that data fits into the

whole case constellation and strategy

too.

Before we end,

can you just give our listeners sense

for how the pricing structure works for

the platform? If people are interested,

how should they go about

inquiring and what does it cost?

So we do the first case

for free. We show you,

not tell you what this looks like.

And then we have firms on our platform

as small as solo shops to as big as

hundreds of attorneys, and we have

pricing that works for every firm.

Ultimately, it's charged back to

the case for most of our clients,

and we give invoices that reflect that

cost so that firms can use the best

tech for their clients. That's

the goal at the end of the day.

So it's nothing crazy.

It costs less than one hour of billable

time and can be the difference between a

$2 million verdict and

a $70 million verdict.

I really do like the way you guys have

developed the pricing structure because

as you point out,

it can be a case cost because it's priced

separately for each individual case.

There is a different price point for

a case review that doesn't have the

full-blown, at least we've been

using that full-blown thing,

but for cases where there's a higher

probability you're going to reject it and

you don't want to be

spending money unnecessarily.

And then once you buy a case,

you can continue to work on it,

it seems like for the life of the case

without incurring additional costs.

So it's worked quite well for us.

I keep waiting for the pricing to go up

because I don't understand how you guys

are making a living over there,

but for right now it's working pretty

well for all of us who are using it.

So appreciate that.

Well, Sarah,

it's so great to have you and thank

you again for developing this wonderful

product.

It's really helped us in our practice

and it's helped a lot of our clients get

to better results.

And hopefully I'm sure that's the

case for lawyers all over the country.

And folks, if you are interested

and want to learn more,

contact Parable and I'm sure they'll

give you the demo and show you what it's

all about.

Thank you for being part of the

Paramble community and inviting us on.

And it sounds like we probably need to

do an in-person training to show you guys

all of the new things that are launching.

I think that's one of the

biggest frustrations we hear

from our customers is it

gets better every week, but now I

need to keep up with all the changes.

So we're actively thinking about how to

make sure we're offering enough training

and support through all this change.

I mean,

you guys send out weekly emails with the

updates and changes and improvements to

the system,

and we actually have designated one

person internally as kind of our parable

guru, and her job is to read those,

to attend trainings occasionally,

and then to report back on

what the new things are.

But actually it would be helpful

to have another firm-wide training.

So maybe I'll reach out and see

if we can arrange for that. Well,

thanks so much for joining.

Thanks for having me.

Did we rise to the challenge

today? If so, tell a friend.

If not, tell us what would make

the podcast more valuable to you.

Thanks for spending your valuable

time with us today. And remember,

when we elevate people

and we elevate practices,

we elevate the profession

produced and powered by LawPods.