September 17, 2025 • 15 mins
I spoke with Karl Schieneman, a Co-Founder of TestMyWitness, which empowers witness preparation using AI. We discussed how legal teams can leverage technology to prepare witnesses to testify, and how that process 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 Carl Hinman, a co-founder of
(00:22):
Test My Witness, which empowers witnesspreparation using artificial intelligence.
Hi Carl, how are you?
I'm doing great, Ari.
Nice to come back to the podcast world.
It's been a break from the legalpodcasting for me since I think 2014.
The death of ESI bites.
I'm looking forward to this conversation.
Tell us about your background andthe genesis of Test My Witness.
(00:46):
Okay.
I've been a serial entrepreneur inthe legal space for about 30 years.
Primarily with Doc Review soldthree companies in the legal space.
Most recent transactionI sold a review right.
And through Inspire review to Haystack id.
I also dabbled in a number ofother things beyond review.
I did the first predictivecoding case that was contested
(01:08):
with Tom Grs Global Aerospace.
I was pretty happy with that and reallyjust tinkered around with problems
that I've seen in law and try tocome up with better ways to do it.
I'm an MBA and an attorney, and thegenesis of test my witness was Carl being
retired, living in Brazil when a friendof his from exchange year in England,
(01:29):
who had been working with AI for fouryears, five years, doing job interviewing
for the French government in England.
And we kept in touch pitchedthis idea about a year ago.
What do you think about using thetechnology for witness prep, and I
thought to myself, what a great idea.
My initial reaction was before Itake my background, which I was
(01:51):
comfortable, I'd ended on a high note.
I wanted to test the technology.
So I said, interestingly, I was involvedin litigation for about three or four
years personally, and I never testified.
So let me take my case and workwith my lawyer and use this tool
before we, we actually go out andtalk to other people about it.
(02:13):
The tool basically is to preparewitnesses to testify, have them make
the mistakes before they're on thestand and get them more comfortable
rather than just have a staticmeeting with an attorney where they're
told, don't guess, don't speculate.
Keep your answers short.
And then they let the witnessgo out and sometimes they do.
Okay.
Sometimes it's train wrap.
(02:36):
How does test my witness work?
It's involves a videotapinga mock testimony.
And so a meeting and can be donejust like it normally does, is
between an attorney and the witness.
But then you videotape, 45minutes to an hour of the attorney
actually be being the other side.
(02:58):
And asking the witness hardquestions, trying to get 'em,
trying to get 'em to make mistakes.
And primarily we use itwith Zoom or teams calls.
You can do it with Google Meet andonce the video is done, you load
it into our system and 10 minuteslater you have a transcript with
a video that shows the witness.
(03:19):
But more than just shows thewitness, there is AI that.
Analyzes the facial expressionsof the witness in real time.
Are they nervous?
Are they, in my case, obstructive?
Are they conductive?
Are they happy?
There's 98 differentemotions that we capture.
And so what the attorney then hasis the ability to grade essentially
(03:43):
the testimony and provide feedbackquestion by question to the witness.
What the witness now has is the abilityto leave the meeting with the attorney,
which could be on Zoom or in person,go home, and then review how they did
with attorney comments and think abouthow they could improve their answers.
(04:08):
It also has some features likeWhatsApp where the attorney has
a comment and , maybe the witnessdoesn't understand the comment and
you can get a little back and forth.
The thing I like the most is it hasthe ability to practice your emotions.
I did it and I was amazed that itturned this German serious guy who came
across as really obstructive the firsttime I did it into a confident person.
(04:33):
When I do demos, I'mlike, I'm not an actor.
There's a dot that I practice onthat's in the app and I smile and I
loosened up my muscles in my face.
We've got 98 different.
Spots that you emote emotion and if you'reonly using a couple of 'em you're missing
the ability to be like a newscaster.
(04:53):
I'm not saying cry on command, I'mtalking about just communicating directly.
There are other emotions, but I wasin the job interviewing business
with staffing for years and, ifyou couldn't talk to the person
interviewing you and connect to them.
Your interview was not a great interview.
There's a baseline of communicating whereyou're connecting, where you have the
(05:15):
right energy, where you're being positive.
And this tool will, on a basiclevel, measure that for all the
witnesses they can practice with that,
how can users trust these results?
First of all, I had the samereaction, as I came outta retirement.
I said, do I wanna come outta retirement?
(05:36):
The beaches are nice herein Brazil, and why do this?
What's the reason?
But I'm a problem solver.
And once I tried it out personally,I said, this is a huge problem.
So on the trust level.
This isn't an AI tool where you'reworried about hallucinations
and things being made up.
(05:56):
This is practice.
This isn't being put in front of a court.
The risk is really not present likeit is for other AI being embarrassed.
The only thing the AI is doing isgiving you a transcript in 10 minutes
that can be annotated by the attorney.
The other thing it's doing isit's showing your expressions.
(06:17):
How do you trust that?
You look at what it says and yougo, I was there saying, I'm nervous.
I was nervous.
There's that there's low risk,but you're also, because it's
not a record, you're creating,it's practice, it's being graded.
And there's also thefact that you were there.
So that, that's how I considerthe trust not being a big issue.
(06:38):
Is it difficult to use?
No, , back when I was trying to getlawyers use predictive coding, we
had to teach statistics and all sortsof things and had mixed results.
It took 10 years maybe.
I think COVID did us a real nice servicein terms of getting people comfortable
to communicating using tools like Zoom.
(06:59):
If I could watch my parents intheir mid eighties navigating Zoom.
There's a button and it's a recordbutton, and you just hit that and
at the end you're done and , itsaves to your computer or to the
cloud, and you just move that fileinto our, you just move a file in.
That's where the AI's creates thetranscript and maps out the emotions.
(07:21):
One other thing I wanna say about trustthat I should have probably mentioned.
AI is not telling the witnessanything other than how a mirror
would tell 'em how they're looking.
It's the attorney who'sproviding the feedback.
It's the attorney who is usingtheir experience as a litigator to
(07:42):
say, when you said that I was gonnahave to cross-examine you to clean
up the record and the naturally anattorney in a deposition would try
to say, no, that was a mistake.
Did you mean to say that, orthey would ask it in their way.
All we're saying is the attorney cancatch those issues before they happen
on live testimony, on video, or awritten deposition, and use those
(08:07):
same tools to say, maybe you wannarethink that answer a little bit.
It was too long.
You gave them more than they asked.
You're guessing, all the thingsyou were telling 'em not to do.
You now have a chance to say, Hey, andthis is what we were talking about in
the meeting not to do, you're doing it.
So we have taught peoplethis way for years in school.
(08:30):
It's called homeworkassignments and quizzes.
You don't go to a lecture, sitthere, say, nod your head and
then take your big final exam.
You get the chance to be graded first.
In education, it's shown, it works.
We're not presenting anything thatI don't think is very controversial
with this idea that feedback is good,more feedback is very good, and they
(08:53):
could be able to do it in 10 minutesbecause of AI makes it possible.
What are the ultimate benefitsof leveraging technology to
prepare witnesses for testimony?
I see a couple benefits.
First of all, the sort of thing thatwe are doing is the sort of thing that
happens in big cases already, exceptinstead of ai, you have a PhD coming
(09:15):
in, or maybe you spend a quarter ofa million dollars in a big case to
hire a mock jury, and they look atthe questioning and they say what they
think of the evidence of the witnesses.
Now we're dropping it downto a few hundred dollars.
So it's almost like a democratizationof litigation tools of being able to
(09:37):
get the feedback the lawyer wants,but more importantly, what the witness
I think wants, because to be thrownto the wolves as a witness is scary.
One of my partners in this tom Thompson,is a trial lawyer and he has used
this three times in cases and he'sthree and oh, and I don't wanna say
(09:59):
it's because of the technology, butit's a good number, so we're gonna
use it better than two and two.
It has been something that allowsa witness to sleep at night is
one of the things he has heardfrom talking to his witnesses.
Allows them to do the best that theycan do to put their best case forward,
are there any disadvantages to creatinga record of witness practice testimony?
(10:23):
I think if you do it the wrong way withoutan attorney present, you run the risk of
creating something that's discoverable.
The record is created on the standunder oath in the deposition.
You don't wanna give an initial,bite of the apple to the other side.
What we suggest is make this work product.
(10:44):
We have on our transcript and everythingthat it is attorney work product done
under the supervision of outside counsel.
Just like when a PhD is hired to assessa motion and testimony and give advice.
So you wanna have an attorney in the mix.
That's one of the reasons wehaven't focused on talking to
companies as much directly because.
(11:07):
For them to sit in front of a computerand answer questions about a case without
an attorney present, I think runs therisk that it becomes discoverable.
But I can't think of many downsidesto getting feedback . Everything
I've done in my career, hasbeen about improving feedback.
The review, software was about testing areviewer before you hire them on review.
(11:30):
Then it was like an Amazon featurewhere you actually had the QC attorney
put notes about the reviewer, how theyrated them from sampling, and these
are the people you wanna work with.
Again, this is feedback.
We did chat in these remote reviews.
Why would you wannacreate a record of that?
I remember getting into an argumentwith an attorney, who said, I don't
(11:51):
wanna create a record but I said, butaren't you trying to do a good job?
Don't you want thereviewers on the same page?
They need the feedback from trialcounsel, when they ask the question,
it needs to go to the team of 30or 50 people at the same time.
Otherwise, people are coding in silosand you're doing everything in, in, in
trying to clean up the review later on.
(12:12):
So feedback just like in those areas.
It's equally important for awitness because the witness
is there on the firing line.
There's no hiding.
Like in a Docker review room wherethere's a QC effort, the witness
messes up your case can be toasted.
How do you see witnesspreparation evolving?
I believe that the technology that we'redoing, this is the most exciting I've
(12:35):
ever been in the field because unlikebeing in Doc Review, where I'm trying to
convince lawyers to hire teams of peoplein the trenches of a case, every single,
litigation or investigation has a witness.
So this is somethingthat everyone can use.
The evolution, is happening because of ai.
(12:56):
This would've been a very hard thing,to convince people to do because lawyers
traditionally don't change very quickly.
But in this current environmentwith ai, , the field is being
forced to grapple with new ways ofdoing things or being left behind.
In my mind, it's not being left behind.
(13:16):
It's the risk of not winning yourcase because your witness messed
up when you could have for a fewhundred dollars worked with the
witness and prevented those mistakes.
And that's a different risk.
For years, lawyers have been able tosay the witness just didn't perform.
And the question I asked was it a badwitness or were they adequately prepared?
(13:39):
And if you're not providing feedback,I'm not sure you can honestly say you did
as much as you could have 10 years ago.
Yes, five year goes, yes.
But now with AI creating a transcriptvery quickly, if you're not providing
this type of feedback you'redoing enough i'm hoping this is an
easier evolution than predictivecoding, which took 10 years to do.
(14:00):
But my view is it's dominoes.
Once a couple big firms fallin the step, and I'm talking
to about 10 of them right now.
Everyone else will say,this is a no-brainer.
Why were you hired to win?
Why didn't predictivecoding go forward quicker?
Let's say people were worried of losingtheir jobs, so I think we can say that
now, and this actually creates morebillable time, not a lot, but it actually
(14:25):
moves the needle in a way where you'redoing real work on a case with your
witness not working on a document thatthey're never gonna see a second draft.
You're actually in the trenches with them.
I think it's very good billable time.
So I don't think we have the same barriersthat you had with predictive coding to
do this, but I will say it is not a spacethat hardly anyone else is in right now.
(14:46):
I think we're the onlypeople just about doing this.
This is Ari Kaplan, speakingwith Carl Hinman, a co-founder of
Test My Witness, which empowerswitness preparation using ai.
Carl, thanks so very much.
Thanks a lot.
Thank you for listening to theReinventing Professionals Podcast.
(15:07):
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 speakingtoday with Carl Hinman, a co-founder of
(00:22):
Test My Witness, which empowers witnesspreparation using artificial intelligence.
Hi Carl, how are you?
I'm doing great, Ari.
Nice to come back to the podcast world.
It's been a break from the legalpodcasting for me since I think 2014.
The death of ESI bites.
I'm looking forward to this conversation.
Tell us about your background andthe genesis of Test My Witness.
(00:46):
Okay.
I've been a serial entrepreneur inthe legal space for about 30 years.
Primarily with Doc Review soldthree companies in the legal space.
Most recent transactionI sold a review right.
And through Inspire review to Haystack id.
I also dabbled in a number ofother things beyond review.
I did the first predictivecoding case that was contested
(01:08):
with Tom Grs Global Aerospace.
I was pretty happy with that and reallyjust tinkered around with problems
that I've seen in law and try tocome up with better ways to do it.
I'm an MBA and an attorney, and thegenesis of test my witness was Carl being
retired, living in Brazil when a friendof his from exchange year in England,
(01:29):
who had been working with AI for fouryears, five years, doing job interviewing
for the French government in England.
And we kept in touch pitchedthis idea about a year ago.
What do you think about using thetechnology for witness prep, and I
thought to myself, what a great idea.
My initial reaction was before Itake my background, which I was
(01:51):
comfortable, I'd ended on a high note.
I wanted to test the technology.
So I said, interestingly, I was involvedin litigation for about three or four
years personally, and I never testified.
So let me take my case and workwith my lawyer and use this tool
before we, we actually go out andtalk to other people about it.
(02:13):
The tool basically is to preparewitnesses to testify, have them make
the mistakes before they're on thestand and get them more comfortable
rather than just have a staticmeeting with an attorney where they're
told, don't guess, don't speculate.
Keep your answers short.
And then they let the witnessgo out and sometimes they do.
Okay.
Sometimes it's train wrap.
(02:36):
How does test my witness work?
It's involves a videotapinga mock testimony.
And so a meeting and can be donejust like it normally does, is
between an attorney and the witness.
But then you videotape, 45minutes to an hour of the attorney
actually be being the other side.
(02:58):
And asking the witness hardquestions, trying to get 'em,
trying to get 'em to make mistakes.
And primarily we use itwith Zoom or teams calls.
You can do it with Google Meet andonce the video is done, you load
it into our system and 10 minuteslater you have a transcript with
a video that shows the witness.
(03:19):
But more than just shows thewitness, there is AI that.
Analyzes the facial expressionsof the witness in real time.
Are they nervous?
Are they, in my case, obstructive?
Are they conductive?
Are they happy?
There's 98 differentemotions that we capture.
And so what the attorney then hasis the ability to grade essentially
(03:43):
the testimony and provide feedbackquestion by question to the witness.
What the witness now has is the abilityto leave the meeting with the attorney,
which could be on Zoom or in person,go home, and then review how they did
with attorney comments and think abouthow they could improve their answers.
(04:08):
It also has some features likeWhatsApp where the attorney has
a comment and , maybe the witnessdoesn't understand the comment and
you can get a little back and forth.
The thing I like the most is it hasthe ability to practice your emotions.
I did it and I was amazed that itturned this German serious guy who came
across as really obstructive the firsttime I did it into a confident person.
(04:33):
When I do demos, I'mlike, I'm not an actor.
There's a dot that I practice onthat's in the app and I smile and I
loosened up my muscles in my face.
We've got 98 different.
Spots that you emote emotion and if you'reonly using a couple of 'em you're missing
the ability to be like a newscaster.
(04:53):
I'm not saying cry on command, I'mtalking about just communicating directly.
There are other emotions, but I wasin the job interviewing business
with staffing for years and, ifyou couldn't talk to the person
interviewing you and connect to them.
Your interview was not a great interview.
There's a baseline of communicating whereyou're connecting, where you have the
(05:15):
right energy, where you're being positive.
And this tool will, on a basiclevel, measure that for all the
witnesses they can practice with that,
how can users trust these results?
First of all, I had the samereaction, as I came outta retirement.
I said, do I wanna come outta retirement?
(05:36):
The beaches are nice herein Brazil, and why do this?
What's the reason?
But I'm a problem solver.
And once I tried it out personally,I said, this is a huge problem.
So on the trust level.
This isn't an AI tool where you'reworried about hallucinations
and things being made up.
(05:56):
This is practice.
This isn't being put in front of a court.
The risk is really not present likeit is for other AI being embarrassed.
The only thing the AI is doing isgiving you a transcript in 10 minutes
that can be annotated by the attorney.
The other thing it's doing isit's showing your expressions.
(06:17):
How do you trust that?
You look at what it says and yougo, I was there saying, I'm nervous.
I was nervous.
There's that there's low risk,but you're also, because it's
not a record, you're creating,it's practice, it's being graded.
And there's also thefact that you were there.
So that, that's how I considerthe trust not being a big issue.
(06:38):
Is it difficult to use?
No, , back when I was trying to getlawyers use predictive coding, we
had to teach statistics and all sortsof things and had mixed results.
It took 10 years maybe.
I think COVID did us a real nice servicein terms of getting people comfortable
to communicating using tools like Zoom.
(06:59):
If I could watch my parents intheir mid eighties navigating Zoom.
There's a button and it's a recordbutton, and you just hit that and
at the end you're done and , itsaves to your computer or to the
cloud, and you just move that fileinto our, you just move a file in.
That's where the AI's creates thetranscript and maps out the emotions.
(07:21):
One other thing I wanna say about trustthat I should have probably mentioned.
AI is not telling the witnessanything other than how a mirror
would tell 'em how they're looking.
It's the attorney who'sproviding the feedback.
It's the attorney who is usingtheir experience as a litigator to
(07:42):
say, when you said that I was gonnahave to cross-examine you to clean
up the record and the naturally anattorney in a deposition would try
to say, no, that was a mistake.
Did you mean to say that, orthey would ask it in their way.
All we're saying is the attorney cancatch those issues before they happen
on live testimony, on video, or awritten deposition, and use those
(08:07):
same tools to say, maybe you wannarethink that answer a little bit.
It was too long.
You gave them more than they asked.
You're guessing, all the thingsyou were telling 'em not to do.
You now have a chance to say, Hey, andthis is what we were talking about in
the meeting not to do, you're doing it.
So we have taught peoplethis way for years in school.
(08:30):
It's called homeworkassignments and quizzes.
You don't go to a lecture, sitthere, say, nod your head and
then take your big final exam.
You get the chance to be graded first.
In education, it's shown, it works.
We're not presenting anything thatI don't think is very controversial
with this idea that feedback is good,more feedback is very good, and they
(08:53):
could be able to do it in 10 minutesbecause of AI makes it possible.
What are the ultimate benefitsof leveraging technology to
prepare witnesses for testimony?
I see a couple benefits.
First of all, the sort of thing thatwe are doing is the sort of thing that
happens in big cases already, exceptinstead of ai, you have a PhD coming
(09:15):
in, or maybe you spend a quarter ofa million dollars in a big case to
hire a mock jury, and they look atthe questioning and they say what they
think of the evidence of the witnesses.
Now we're dropping it downto a few hundred dollars.
So it's almost like a democratizationof litigation tools of being able to
(09:37):
get the feedback the lawyer wants,but more importantly, what the witness
I think wants, because to be thrownto the wolves as a witness is scary.
One of my partners in this tom Thompson,is a trial lawyer and he has used
this three times in cases and he'sthree and oh, and I don't wanna say
(09:59):
it's because of the technology, butit's a good number, so we're gonna
use it better than two and two.
It has been something that allowsa witness to sleep at night is
one of the things he has heardfrom talking to his witnesses.
Allows them to do the best that theycan do to put their best case forward,
are there any disadvantages to creatinga record of witness practice testimony?
(10:23):
I think if you do it the wrong way withoutan attorney present, you run the risk of
creating something that's discoverable.
The record is created on the standunder oath in the deposition.
You don't wanna give an initial,bite of the apple to the other side.
What we suggest is make this work product.
(10:44):
We have on our transcript and everythingthat it is attorney work product done
under the supervision of outside counsel.
Just like when a PhD is hired to assessa motion and testimony and give advice.
So you wanna have an attorney in the mix.
That's one of the reasons wehaven't focused on talking to
companies as much directly because.
(11:07):
For them to sit in front of a computerand answer questions about a case without
an attorney present, I think runs therisk that it becomes discoverable.
But I can't think of many downsidesto getting feedback . Everything
I've done in my career, hasbeen about improving feedback.
The review, software was about testing areviewer before you hire them on review.
(11:30):
Then it was like an Amazon featurewhere you actually had the QC attorney
put notes about the reviewer, how theyrated them from sampling, and these
are the people you wanna work with.
Again, this is feedback.
We did chat in these remote reviews.
Why would you wannacreate a record of that?
I remember getting into an argumentwith an attorney, who said, I don't
(11:51):
wanna create a record but I said, butaren't you trying to do a good job?
Don't you want thereviewers on the same page?
They need the feedback from trialcounsel, when they ask the question,
it needs to go to the team of 30or 50 people at the same time.
Otherwise, people are coding in silosand you're doing everything in, in, in
trying to clean up the review later on.
(12:12):
So feedback just like in those areas.
It's equally important for awitness because the witness
is there on the firing line.
There's no hiding.
Like in a Docker review room wherethere's a QC effort, the witness
messes up your case can be toasted.
How do you see witnesspreparation evolving?
I believe that the technology that we'redoing, this is the most exciting I've
(12:35):
ever been in the field because unlikebeing in Doc Review, where I'm trying to
convince lawyers to hire teams of peoplein the trenches of a case, every single,
litigation or investigation has a witness.
So this is somethingthat everyone can use.
The evolution, is happening because of ai.
(12:56):
This would've been a very hard thing,to convince people to do because lawyers
traditionally don't change very quickly.
But in this current environmentwith ai, , the field is being
forced to grapple with new ways ofdoing things or being left behind.
In my mind, it's not being left behind.
(13:16):
It's the risk of not winning yourcase because your witness messed
up when you could have for a fewhundred dollars worked with the
witness and prevented those mistakes.
And that's a different risk.
For years, lawyers have been able tosay the witness just didn't perform.
And the question I asked was it a badwitness or were they adequately prepared?
(13:39):
And if you're not providing feedback,I'm not sure you can honestly say you did
as much as you could have 10 years ago.
Yes, five year goes, yes.
But now with AI creating a transcriptvery quickly, if you're not providing
this type of feedback you'redoing enough i'm hoping this is an
easier evolution than predictivecoding, which took 10 years to do.
(14:00):
But my view is it's dominoes.
Once a couple big firms fallin the step, and I'm talking
to about 10 of them right now.
Everyone else will say,this is a no-brainer.
Why were you hired to win?
Why didn't predictivecoding go forward quicker?
Let's say people were worried of losingtheir jobs, so I think we can say that
now, and this actually creates morebillable time, not a lot, but it actually
(14:25):
moves the needle in a way where you'redoing real work on a case with your
witness not working on a document thatthey're never gonna see a second draft.
You're actually in the trenches with them.
I think it's very good billable time.
So I don't think we have the same barriersthat you had with predictive coding to
do this, but I will say it is not a spacethat hardly anyone else is in right now.
(14:46):
I think we're the onlypeople just about doing this.
This is Ari Kaplan, speakingwith Carl Hinman, a co-founder of
Test My Witness, which empowerswitness preparation using ai.
Carl, thanks so very much.
Thanks a lot.
Thank you for listening to theReinventing Professionals Podcast.
(15:07):
Visit reinventing professionals.com orari kaplan advisors.com to learn more.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
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
CrimeLess: Hillbilly Heist
It’s 1996 in rural North Carolina, and an oddball crew makes history when they pull off America’s third largest cash heist. But it’s all downhill from there. Join host Johnny Knoxville as he unspools a wild and woolly tale about a group of regular ‘ol folks who risked it all for a chance at a better life. CrimeLess: Hillbilly Heist answers the question: what would you do with 17.3 million dollars? The answer includes diamond rings, mansions, velvet Elvis paintings, plus a run for the border, murder-for-hire-plots, and FBI busts.