From Wrenches to Algorithms: How AI is Redefining Auto Repair
All right, rolling.
This is it.
Exciting.
Texas, Houston, Texas.
I've never been here.
You haven't either?
No, no.
Flew in.
Fourth largest city, I heard.
I was doing a little culture
review on the way in here,
but it's kind of a slept on city,
I feel like.
The only thing I know about
it is like space, Houston.
I guess, oh, Houston, we have a problem,
huh?
That's true.
Something's wrong.
Houston, we definitely have a problem.
That's what's going on right now.
That's the only thing I know
about Houston.
It's like I asked a lot of
guys that would come here like, oh,
you're from Texas.
Like, oh,
you're like right around the corner then.
But you forget like how big
Texas is like you're not
right around the corner.
I mean, I'm from Fargo, North Dakota.
So I'm like somewhere in Canada.
This is,
this is foreign country to me down
here and down here in Texas.
You're like way away.
Yeah.
I'm California.
So four hour drive, three hour flight,
two hour time difference.
So we're in California, uh, past Robles.
So right.
California coast.
Okay.
Right in the middle.
About ten, fifteen minutes inland.
I spent pretty much the past year in LA.
Pretty comfortable.
I moved there right for those fires.
Oh, perfect.
That's not a good time to be there.
No, it was terrible timing for sure.
Well, tell me about yourself.
Yeah.
Name's Josh Oberlander.
Just have an AI company here
in the automotive space and
we're just getting the ball rolling.
So we have only been around
for about nine months now.
Wow.
And so it's gone super exciting.
Been in the industry for about three years,
but really been only doing
this company called Detect Auto.
I'm one of the co-founders
with two other boys.
We've been doing that for
about nine months and been loving it.
Yeah.
I guess in the world of AI, that's what,
ten years?
Yeah, a long time.
We're already behind the power curve.
Yeah, it evolves so fast.
So there's a lot of AI stuff
popping up right now,
especially in the auto industry.
A lot of negative feedback,
a lot of negative
connotation when it comes to AI.
I think not just in our industry,
but in general, in the whole world.
Everyone's like, oh,
they're going to take our jobs.
It's what everyone's so worried about.
I don't know if the touch
points are ever going to go away,
but what is exactly that
you're establishing in that industry?
So, I mean, I would agree.
First of all, I do think it's scary.
Even me, somebody who's building AI,
I'm like half the time
worried about my own job.
Yeah.
Where my own job will be in five years.
And I'm the one,
we're the ones creating
some of this stuff.
So.
Um, I would say, you know, it's,
it all comes down to just
making somebody more efficient, right.
And, and whatever process.
And so one thing that we do is, um,
we make you more efficient
on your software.
So, uh, when it comes to typing,
when it comes to manually
researching vehicles,
when it comes to like some
monotonous task that you
have to do all day,
every day over and over and over again,
um, that's really where we shine through.
And, um,
We have tools, we call tools,
but we don't do exactly one thing.
We find areas that are
problematic and and people
spend way too much time on
and that's kind of so we're
trying to just not
obviously it sounds to me
like there's a lot more of
a human element that's
involved to make it work is
that a fair assumption it's
because without knowing
where the inefficiencies so
the greatest thing about ai
is it makes processes
easier when you can put in
exactly what you want it to
do right um if you wanted
ai to go look out and say
on a vehicle intake
see that there's a car seat and there's a,
um, what's, what's your football team,
right?
Like whatever that,
whatever the football
team's on the back bumper, right?
Yeah.
Um, American football, not soccer, right?
Yeah.
But anyway, Vikings, I guess would be our,
okay.
So, so they have a green basic or not.
I mean, to do that.
Yeah.
So they had a big cheese
head sticker on the back.
Whatever it is,
that's not something AI is
ever going to be able to do.
They're never going to be
able to pick up on nuances
of the customer interaction,
in my opinion.
There's always going to be something.
You can call,
and we've all seen these AI
videos on social media.
They call you at three in the morning,
right?
And they answer all the
customers' questions, right?
Mm-hmm.
but they don't build a relationship.
Yeah.
Right.
Yeah.
And it's just, it's so close,
but it's missing that.
And I was like, what is it missing?
Like you see all these videos.
It's like, well,
it's missing the human element.
Like it's missing the fact
that it can bond with you.
Make it, make a, you know, Hey,
you're from up in Canada.
Hey, I got,
I got some buddies up in Canada.
You ever heard of Jeff Compton?
He does a Jada mechanic podcast.
He's up by Canada, right?
Like it's never going to
make those kinds of
connections to see if we
have something we can relate to.
You know what I mean?
Yeah.
So I guess my point to get
off my tangent was, um,
With your software that you're creating,
it sounds like you're going
to need someone to direct
that AI to then improve the process.
I guess I got to start from square one.
What does it do?
How does someone get your program,
your process, and stay one in the shop?
What does that look like?
Yeah, but I would totally agree.
AI, especially right now, building trust,
that human factor,
it's what it's incapable of doing.
But one of the tools that we have is...
let's say for example, a car comes in,
what we do is we research
that car and tell you
exactly what maintenance is
due on that vehicle.
So typically the monotonous
task would be you pull up
all of your history,
anything you guys have done in the shop,
and you're looking through
all of the history that you
guys have done,
you look through the Carfax,
you compare it to the
maintenance interval,
you try and do the math back and forth of,
hey, this was done there, is it due now?
And you're doing this on fifteen items.
you know, that you want to look for.
So our,
one of our tools just does that
automatically.
So as soon as the RO is created,
you just click a button.
There's a full report on
exactly what is due on that
vehicle in terms of time
and mileage for that
customer based on its history.
So super simple tool.
So is that pulling the
history from prior reports
or from Carfax or from all of the above?
From all the above.
Yep.
Yep.
So there would have to be a
database for it to pull from.
In order for it to be effective,
obviously.
Yeah.
So that's why we integrate
with other shop management softwares.
So right now we're working
with Techmetric and Shopware.
So you still have to have
that shop management
software for us to be
pulling that information
from all that history that
you guys have previously or
the shop has previously done.
And that's what we're looking for.
Yeah.
I feel like that is kind of
a missing link.
as a vehicle comes into the shop, you know,
one of our protocols is to
have the service advisors
or the technicians look in
to see what the prior
service history was that's
in the reports or,
simply go through the
vehicle and see if there's
anything that's been fixed recently,
right?
So it is a pain point.
I could see needing that.
I could see a lot of friction saying, well,
that's what the advisors are paid for,
right?
Like that's what they're there for, right?
Why are you having this new
software when you already
have a body that can just
simply pull on reports and
being familiar with shopware,
it is quite an easy report to pull up.
Does it take thirty seconds?
Yeah.
Yeah.
But what's what's what's
your thought on on on that
kind of comment?
Yeah.
I mean, yeah.
obviously somebody can do it.
Like it's, uh, it's,
it would be part of their role,
but if it's that the shop gets super busy,
um,
and it's chaotic and you didn't have
time to do it before the
customers in there.
Now,
now your front of house is all chaos
with five people in there.
And, um,
it seems to always be one of the
first tasks that gets pushed to the side.
They sell the,
the easier work to sell is
the safety concern.
Hey, my technician saw this.
This is a problem.
You need to do this.
Um, and yeah,
That other maintenance work
sometimes is a little bit harder to sell.
It just it seems to be one
of the tasks that falls off.
Yeah, I can see.
I can see that you being
useful so that during the
inspection process, you're not.
you know,
recommending something that's
already been done or maybe
does cause it's under warranty or,
or what have you, right?
Like there's plenty of parts that fail.
And so you want to be proactive and say,
Hey, we noticed your whatever's leaking.
And we just did that about a year ago.
So we're going to go ahead
and warranty that for you.
And it takes a lot of that,
that pain or that pressure, that,
that uncomfortable phone
call where you're like, Hey, Mrs. Jones,
we valve cover gas is really good.
Well, you guys just did that a year ago.
Like, Oh,
that's just putting your foot in
your mouth.
Yeah,
and especially if you have a fluid
warranty or something like that,
it's pretty difficult
sometimes to keep people on track.
Maybe, you know,
they actually went over
their thirty K warranty.
Sure.
Our system would be very
good at just making sure
you catch it before that goes over.
Yeah, that makes sense.
I could I could see that being useful.
I guess what my question was
targeting is who the shops
that you're looking to to get into.
Right.
So I feel like someone like one technician,
no advisors, just the owner in tech.
I feel like maybe that's not
your target market, right?
And you're kind of looking
for the multi-shop or maybe
something in between.
What's,
do you have like a client avatar
that you have created?
Yeah, I would say, um,
definitely most of our
tools target the service
advisor and helping them
make their day to day a lot easier.
So, um,
And it would also be good
for getting a team.
I would say a lot of our tools,
if you have a super
experienced front of house,
they're already going to know what to do.
You have the procedure in there.
They, they're probably, you know,
can do everything our tool does.
But it's really good for a newer team.
So somebody who, you know,
is getting comfortable or
doesn't know the process already,
we could just completely do it for them.
So for example, just for example,
another one of our tools
helps them get more
information out of the
customer and by asking good questions.
So basically customer walks
in the door and they say, hey,
I have this weird noise
coming from my car.
If the technician just got
that and they didn't do
like a little research and
asking good questions to
the customer and the technician just got,
hey, there's a random noise here.
They could die out the wrong
noise or they could spend a
lot more time dying.
What our system does is it
gives them technical, not technical.
It gives them easy questions
to ask the customer to get
more information out of
them so that the technician
spend less time dying.
So it's really just good for
a newer team member,
somebody that isn't technical,
that doesn't have that experience,
but maybe they're a great salesperson.
So it could open up the door
to you being able to hire more people,
a more diverse group.
And that's what,
I think that's what a lot of our tools do.
They just empower people and
give it a wider range is
able to take on that role.
I feel like you're kind of
dumbing down AI in a sense.
If that resonates.
Whereas, you know,
we have open AI and chat GBT and sure,
you can add stuff in there, right?
Like you can use it to your
advantage if you know how
to script or you know how
to prompt it correctly, right?
And it sounds like what
you're doing is designing
something that makes it a little bit more
palatable right and you're
designing it so like don't
worry about the prompts we
got all that taken care of
here just use this and this
will it gives you all the
power of that without going
outside of your lane or
like outside of your bounds
yeah does that sound kind
of like what's going on
I think that's a perfect
explanation of it.
That's what it sounds like.
So two things.
We try and make it super easy,
as easy as possible.
There's only one path
forward for you to go.
And number two,
we build it in the place
they're most comfortable,
the shop management
software that they use every day.
They don't have to go to another site.
It works directly inside
where their team is used to
working all the time.
So they don't have to leave,
waste time going somewhere.
They don't have to copy and paste.
It interacts all natively right inside
And this sounds like
something that's going to
evolve with that shop, correct?
So you don't have a fresh, I mean,
the chat,
like the chat for O is good
where you can keep different groups,
right?
So you have different,
what do they call them?
Bots,
different bots for different subjects.
So this sounds like it's
probably going to keep
integrating with the shop.
So you're not having to start fresh,
right?
So you can say,
this is our branding package, right?
This is how we communicate.
This is what we say.
These are our taglines, right?
Um, I like to say we're,
we're shifting gears, auto repair.
So we're shifting expectations.
So I feel like that's
something I can incorporate.
And then it would always use that as a,
as a plug, I guess.
Yeah.
You can definitely build out
your own formats.
You can build out your own questions.
I mean, most people have, you know, good,
especially experienced front of house.
They have questions.
They know that work every
time somebody comes in with
a check engine light.
Right.
So you can train it,
you can teach it or you'll just learn.
So yes.
Okay.
Sounds like there's a lot going on,
a lot more than the surface
level stuff we're talking
with when it comes to this program.
Because it kind of sounded
like a little bit of just
helping take on the smaller tasks,
but it definitely sounds
like you can get it a lot
deeper into your shop as well.
Yeah.
Now,
what about the fear base where they're
like, oh,
now your whole shop just ran by AI,
right?
Yeah.
Yeah.
I mean, um, I,
I personally think there
still needs to be,
there's going to be a
person in the loop no matter what this,
the AI stuff, at least my knowledge,
I use it every day, all the time.
It needs to be checked.
It hallucinates information.
It comes up with stuff that they're like,
I, I even prompted you not to say that.
And you're still,
it's going like further
down a rabbit hole.
And so I think, um,
I think that there's still
going to be a person there in the loop.
And it's just maybe you
won't need three service advisors.
It'll just be two.
I see.
Yeah,
there's a lot of that where advisors
are going to be fired or
advisors are going to be
replaced by AI in a sense.
And the way I look at it is
what's it been about?
Ten years now?
Twelve years?
Remote advising?
Yeah.
Right.
I don't know if you if
you're familiar with that, but yeah,
there's call centers that
you can hire for your shop.
So all your all your calls
are then forwarded to a
third party out in
Indonesia or somewhere along those lines.
Like that's where those call centers are.
Right.
People are getting paid two bucks an hour.
Right.
So you have someone that
answers a phone there and
then redirects it once they
go through the script.
Right.
And that's been going on for a long time.
And I just don't know where
this whole freaking out
about chat GBT coming in.
Like, have you not been paying attention?
Do you know what a lot of
these MSOs are doing right
now with call centers?
Like, it's kind of crazy, right?
Like,
I'd rather have chat than pay someone,
you know, in a third world country.
Like, that's wild, man.
That's the last thing I wanted.
Like,
we all have help from Verizon or
somebody like that.
We've all had those conversations, right?
And you can tell pretty quick.
Yeah.
You can tell pretty quickly.
These companies are like,
that ain't working anymore.
Cause now you call and you
actually talk to someone in
the U S you're like, Oh, there we go.
Yeah.
Someone that understands my
freaking language.
All of the above though.
Again, it goes back to losing trust.
I think without having that
person in person
communication and building
a relationship with them overall,
all of it just lose trust.
So I don't know exactly
where that will head, but it,
that doesn't seem like a good thing.
I think the point here is
it's not black or white.
And I think everyone likes
those two colors so much, right?
Like it's either that or it's that, right?
And they forget about all
the gray in between, right?
I think here, seeing this,
the auto shop answers and
the coaching and seeing
that there's a lot of shades of gray,
right?
Like everyone wants to
standardize and an SOP.
And I feel like the secret kind of is...
in between there right is
being able to do the the
broad strokes of gray and
being able to run your
business based on that
right um it's it's
difficult because it's so
easy to put someone in a
box right or like say this
is exactly how it's going
to be every single time and
then you can walk away and
it's supposed to run like
that forever but we're in
the like we're in the
customer relation business
right like yeah you just
can't yeah no I totally
agree with that yeah
And I don't know if AI is
ever going to get to that
point where it's going to
differentiate between the two, you know,
and I've used and I'm sure you have to.
And sometimes you'll get a
little bit different
response on the same exact question.
And that's kind of strange.
And it's like, is that better or worse?
Right.
Is it is it trying to like
paint between the lines or
like what's what's the deal with that?
And I've seen a lot of prompts, too,
where you put something in that says,
let me get back to you.
And then you're like,
is this thing thinking right now?
I don't know if you had a,
have you had a problem like that yet?
I mean, I have it all the time.
I mean, we, I could,
a customer could come in and they have,
again,
like that noise complaint and it
could give me five perfect questions.
And the next time it's like
five other questions.
Right.
Like what,
what was wrong with the five
questions that I perfectly
liked last time?
That's true.
You know, I, I do think it's,
doesn't like to give the same information,
which is very weird.
It does like to push out new
stuff every time.
What about the thinking model?
Have you...
Witness that or do you know
anything about like the
thinking model when it comes?
Sometimes you add something
in there and it says,
give me twenty four hours
and I'll get back to you.
I have not seen it.
You haven't seen that?
No.
That's like the newest
iteration I've seen is that
you put something.
Let me get back to twenty
four hours and then you
wait twenty four hours and then it's like,
wait a second.
Like, is this thinking right now?
Like, is it coming up with an answer?
Yeah.
What kind of questions are
you asking that?
Oh, I don't remember.
I don't remember.
that thing normally solves
anything I have in two seconds.
Let me give you some, let me,
let me think about this or not think,
but it says,
let me get back to the twenty four.
Like what the hell?
I've never seen that before.
Yeah.
And I did some research and it was like,
is there, is there a thinking bot?
Right.
I'm like,
is this thing evolving to the
point where it's thinking?
I don't think so.
I don't,
I don't know if you knew anything
about that.
So I wanted to ask you, I did not,
I did not.
You probably work a lot
closer with it than I do.
And I'm, I just,
there's been a few times
where I've kind of broke it.
And part of it's with the podcast,
you know,
I'll transcript the product podcast.
And then it's like,
error, you know,
like a one hour transcript and it's like,
nope.
I'm like, oh, really?
I think what that is,
how crazy good it's gotten
with all the like videos,
all the pictures.
So basically in the past like year,
like last year,
it couldn't look at a PDF
and like properly analyze it very well.
That's true.
And now this year you can
have it make whatever image you want.
Yeah.
And it still is pretty goofy sometimes,
but like compared to like
Two years ago.
What's with the words?
Like you say like, Hey, I want this.
And then it's like,
it doesn't spell the words
in like banners and stuff.
Yeah.
You know what I mean?
Like, how do you, what?
You don't see that?
Like it's, it's, it's bizarre.
It is.
It is weird.
Right.
Why it doesn't get that.
But like the,
one of our tools it's able to
do is it can analyze photos
that the technicians take.
That's what I was going to ask next.
Yeah.
During their DVI.
So they can just go through,
snap some photos, you know,
I'm taking photos of, you know,
I don't know.
their brakes or their filters or, uh,
their cracked windshield or
dents on the outside.
And they don't even need to
give it anything.
It can analyze the photo,
which it does has been
doing like really good.
We just launched this
product about a month ago and, uh,
can write up a finding for them.
It saves them time.
So I think, you know, one of our,
at least for me,
I hate typing and I know a
lot of technicians, um, they don't know,
they don't know how they
don't spell good.
No, I'm just kidding.
Or they use like the voice to text,
which is super efficient.
I love that.
But it comes up with a lot of mistakes.
And so like one of our tools
cleans up all that and just
makes it super professional
and cleans up your grammar
and your spelling,
like simple things like that.
They leave the talk to text
on and then you get this
whole conversation about
something you definitely
didn't want to know about.
Like, hey, bud,
you left your talk to text
on for like about five minutes.
Yeah, under your brain inspection.
It's like a huge paragraph.
Yeah.
So but I mean, it's pretty cool.
Like you could just like give a color.
Let's say you're like you're
doing the inspection.
You mark your brakes as
yellow and that they're
five millimeter and it can
write up a whole statement
looking super professional saying like,
hey, you need to monitor these.
They're coming to like, you know,
just a type of or the
thickness that you need to
start monitoring.
Sure.
Sure.
And that sounds like you
could probably put can a can summaries.
So that way it's not a
different thing for every, you know,
you want the same
experience for the customer every time.
So once it does the one time,
it's just going to use that every time.
Yeah.
You know, that makes sense.
I can see that being super helpful with,
with the inspections and
trying to stay on top of that.
So yeah.
What's your biggest feedback
that you've gotten this
year with implementing it?
with specifically the like
photos and stuff?
Just in general, have you had any,
like what's your,
any hopefully positive feedback,
but any feedback in general
that you want to share
about shops that have
implemented this and what
they've seen the difference in?
Yeah, I mean,
definitely selling more gravy work.
That's our overall goal.
I think people have been super excited.
I think somebody who wasn't
in the industry, me,
I initially came in and
heard how hiring is one of
the biggest problems.
I know technicians and even
service advisors,
finding good people to work up front.
And I think our tools have
been able to make
you know, uh,
give you a more diverse crowd
that you can hire from.
So I think people have been super excited.
Uh, I have,
we have almost had no negative feedback,
um, over the past couple of months.
And, um, yeah,
I think it's all been pretty good.
That's awesome.
Yeah.
What's the, um, how does, so with the,
with the program,
you have to have a certain
management program, obviously to, to,
to integrate.
Um,
what does that process look like to set
up and to get someone on
board if they're interested
in wanting to do it?
Yeah, it's super simple.
You download a simple
extension that goes onto
your Google Chrome browser
or your Edge browser or
whatever internet browser you're using.
And it works in about two seconds.
Okay.
So super simple setup.
No onboarding process.
I mean, if you want to do maintenance,
we can onboard.
You track which items you want,
get those intervals in there.
Are you using OEM stuff?
There's a lot of specifications,
but in order to actually
just start using software
with some of these tools
like the asking good
questions and the text
write-up and cleanups,
it works right away.
Okay, nice.
So super simple.
And is the program slowly evolving?
So as it goes on,
are you going to be able to
use that AI model to...
be more established for new
shops or is every
integration going to kind
of start from square one?
Does that make sense?
Yeah, I think, um,
so we're kind of learning
from everybody and, uh,
I think the baseline model
is adequate for, you know, like you can,
you can,
it can learn and get better from there,
but, uh,
Is the plan to roll out more
advanced versions as they
learn from the other shops
and then you can sell the
advanced version at a different rate?
Like, hey,
this has already got all this
learning capability already done.
It's, you know, and then you could say,
here's your, here's your, your assistant.
I mean, yeah, I would say, uh,
Definitely,
especially when it comes to
writing up ROs, looking at your notes,
how you write things.
I would say the customer
concerns and asking
technical questions up front,
that's pretty simple.
But when it comes to helping
diag a situation or things like that,
that's kind of where I
think it needs a lot of learning.
Oh, yeah, I can see that.
I mean,
as far as having shops run it in
their shop for a certain amount of time,
and it's going to learn their process,
right?
Is that like a proprietary bot now?
Or is that software still
going to be technically
yours where you can use
that to then implement it
into another shop?
Does that make sense?
Kind of.
I would say... And like a learning model,
you'll have a certain DVI
that you want to share.
So let's say there's like
DVI programs out there, right?
And so over time,
they'll build a DVI and then they can say,
hey,
do you want to make this inspection
shareable or non-shareable, right?
And other shops say, oh,
that's a perfect inspection model.
I'll use that so I don't build my own.
Yes.
Is that something that
you're going to develop or
you're going to be able to...
Yes, I see what you're saying.
Yes.
Yeah, it's super easy to as soon as we,
you know, build that out.
And or if you have formats,
it would be very easy to
share or not share.
I guess my fear is like you have a chat,
you know,
bought with your shop for four
or five years.
It learns and perfects everything.
And then all of a sudden it's like, oh,
we'll just share that AI
model with with every other shop.
Right.
Yeah.
Yeah.
I can see some people being
pretty upset about
something like that happening.
Yeah.
We definitely don't want to
be giving out everybody's, uh, you know,
secret sauce.
Yeah.
I mean, not that there really is,
but some people really think there is.
So, you know, for sure.
Everybody kind of stinks.
Everyone's like,
What you do at your shop
doesn't work at your neighbor's shop,
right?
So it's not going to work
across the country.
It's not going to work in the same state.
It's everyone's got to do it.
But for some reason, everyone thinks, oh,
you're going to steal my stuff.
Yeah.
You could have the perfect
DBI process built out that
works insanely well in your shop.
But that doesn't mean a
different shop that
implements that is going to
have the process.
The shop down the road, right?
It's still not going to do it.
Yeah.
That makes sense.
Yeah.
Well, right on.
Anybody wants to get in touch with you.
How do they, what do they, how do we,
all that good stuff.
Yeah.
We're called detect auto and
you could just detect.
Yeah.
Detect auto.
There it is.
Two words right here.
So, um, yeah, just check out our website,
www.detectauto.com.
Um, we got a free trial, all of our stuff.
You can try it out for free,
see which products you like.
We've got four cool tools, um,
that anybody can use.
Is it contract month to month?
Just month to month.
So that thirty day free free
trial and then it's no contract.
No set up fees.
No set up fees.
You just roll.
Yeah.
So we really we I mean,
we're a new company just
trying to get people see
how they like all these these new tools.
So we're trying to get as
many people on the platform
to you and give us feedback.
Really been just trying to
grow over the past couple of months.
We just started with one tool, like I said,
eight months ago.
We're up to four now.
built solely based on feedback from,
from customers.
So it's a good, good spot to get in and,
and, uh, help,
help you grow and help all
shops grow and be on the
ground floor or something.
Uh,
you're still working with the
co-founders.
So like I'm one of the ones
that are on boards, everybody.
So, um,
and then I turn around and my
roommate is the CTO who's
building everything in the
next room over.
So it's a pretty short cycle there.
The customer service is a
lot better than calling Indonesia.
Yes, I hope so.
Or else I'm doing something wrong,
for sure.
Well, right on.
Well, thanks for coming on and sharing,
man.
Pleasure.
Yeah, it was a fun time.
Yeah.
Creators and Guests
