March 13, 2025 • 33 mins
For the March 2025 episode, co-hosts Ted Stank and Tom Goldsby are joined by guest host Thomas Deakins, managing director of the Global Supply Chain Institute, and Jason Murray, CEO and co-founder at Shipium to explore how data-driven decision-making fuels modern supply chain success.
Their in-depth conversation touches on the power of predictive technology and Agentic AI, the complexities of integrating products and software, and the critical role of data in solving today’s most pressing supply chain challenges.
With 19 years at Amazon in VP roles over retail systems and services and forecasting and supply chain, Murray brings deep expertise to the discussion. At Shipium, which he founded after leaving Amazon in 2019, he’s focused on bridging the gap between pre-purchase delivery promises and post-purchase fulfillment—ensuring better outcomes for both retailers and consumers.
Listen in as our hosts also break down the strategic use of tariffs in trade negotiations, shifting consumer sentiment amid recession fears, and how the stock market reacts to supply chain uncertainty and policy shifts.
The episode was recorded virtually on March 6, 2025.
Related links:
- Tom Goldsby shares insights on the effects of a trade war on supply chains
- Consumer outlook sinks as recession fears take hold
- Stock market continues its slide after Trump ups the ante in trade war
- Trump doubles planned tariffs on Canadian steel and aluminum
- U.S. employers add 151,000 jobs in February, unemploment up to 4.1%
- Jack Daniel’s producer claims Canada’s empty shelves ‘worse than tariffs’
- Register for the Supply Chain Forum, April 8–10 in Knoxville
- Join the Advanced Supply Chain Collaborative to explore advanced concepts in SCM with experts
- Read our latest white paper, “Unlocking the Potential of the Gen XYZ Labor Force”
- Become a Supply Chain Forum member
- Subscribe to GSCI’s monthly newsletter
- Read the latest news and insights from GSCI
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Intro & Outro (00:00):
Welcome to the Tennessee on Supply Chain
Management podcast.
Listen in as co-hosts Ted Stankand Tom Goldsby set sail into
the world of end-to-end supplychain management, diving deep
into today's most relevantbusiness topics.
They'll share insights inpressing industry issues and
tackle the challenges keepingsupply chain professionals up at
night.
If you're enjoying the ride,download and subscribe to
(00:22):
Tennessee on Supply ChainManagement on your favorite
podcast platform now.
Ted Stank (00:28):
Hello everyone and welcome to Tennessee on Supply
Chain Management.
This is our sixth episode ofseason three.
A lot of crazy stuff going on.
As we sit here recording, it'searly March 2025, six weeks into
the new Trump 2.0administration, and if anybody
is not on the far side of themoon, you'll know that there is
(00:50):
just a lot of change going on.
Much of that change isimpacting the supply chain.
We'll touch on some of thattoday and then we'll talk about
how do we stay flexible?
What are tools we can use totry to help us?
In this time, we'll bring onour guest, jason Murray, to talk
about AI applications.
In this time, we'll bring onour guest, jason Murray, to talk
about AI applications in supplychain.
Before we get into that, let'sintroduce my co-host, great
(01:11):
friend, tom Goldsby.
Tom Goldsby (01:12):
Hello, ted, great to be back with you.
Our loyal listeners may haverealized that I was missing on
the last episode, butfortunately we had Josh King,
the Director of Marketing andOperations for the Global Supply
Chain Institute here at UT, dosuch a great job on the last
podcast.
That said, I'm glad to be backin the mix, but hey, we've got a
(01:34):
real full room today, so let'sget around to introducing some
of the other folks.
Ted Stank (01:39):
Good deal.
One of the other people joiningus as co-host is our own Thomas
Deakins.
Thomas is Managing Director ofour own Thomas Deakins Thomas is
managing director of our GlobalSupply Chain Institute.
He's also a UT supply chainmanagement undergrad and MBA
former vice president of Project44, and also worked with
companies like Oracle, G-Log,Trimble, Mavenwire, MercuryGate,
(02:00):
CHEP, SLI Consulting and JBHunt, so pretty much covered the
spectrum of supply chaincompanies, Thomas, welcome.
Thomas Deakins (02:08):
Hey, happy to be here and honored to be part of
the show, so looking forward toit.
Ted Stank (02:14):
Great Also like to welcome our guest, Jason Murray.
Jason is co-founder and CEO atShippium after a 19-year career
and ended up as a vice presidentwith Amazon.
Jason's been with us as apartner on Global Supply Chain
Institute for a while now.
He spoke to our executiveadvisory board last year and
just blew us all away with whathe's doing at Chipium and the
(02:35):
promise of AI in helping us inuncertain times.
Again, as I hedged that early,if you don't think that we're
living in uncertain times, thenyou probably have been on the
far side of the moon, Jason,welcome.
Jason Murray (02:49):
Hey, great to be here.
Nice to see you guys, I'mexcited.
Ted Stank (02:51):
Hey, I mentioned a little while ago that we had our
executive advisory board.
We recorded our podcast fromthat executive advisory board
last month.
A lot of great discussion.
One of the big topics was howdo we apply AI in the supply
chain?
I think it's one of thoseissues that we all know it's
coming.
Everybody is trying it indifferent ways.
I'm not sure yet that there aresome great use cases, although
(03:14):
hopefully Jason will be able toshed some light on that.
We had about 30 of our advisoryboard members present, had some
great presentations on some ofthe projects we're working on in
the Advanced Supply Chain,collaborative around planning,
around how do we improveretention of Gen Y and Gen Z
labor, particularlymanufacturing and operations.
(03:35):
So a lot of really good stuff.
It was great to have them allthere.
Thomas, we also have somethingbig coming up in the second week
in April.
You want to talk a little bitabout Forum.
Thomas Deakins (03:44):
Yeah, thanks, ted.
April 8th through 10th we willhave our Supply Chain Forum.
It'll be the one for the spring.
Downtown Knoxville Marriottagain dates April 8th through
the 10th.
So if you haven't registered,please do so.
Look forward to seeingeverybody.
We've got a great lineup ofcontent.
Two of the guest speakers I'llhighlight real quick.
(04:06):
One will be Marianne Wanamaker,who's the Dean of the Baker
School for Policy and formereconomist at the White House in
the Trump 1.0.
And she will give us a state ofthe economy, so we're so
excited to hear about that.
And then the second one is thenew voice of the Vols, mike
Keith.
Mike used to be the voice ofthe Titans.
(04:28):
It'll be great to hear from himand also just the ties that he
has back to our department.
Ted Stank (04:33):
I think Mike came over to Tennessee because the
Vols are looking a lot betterthan the Titans these days.
That's not hard.
The Titans used to playfootball, but I'm not sure
anymore.
This is said as a person who isa New York Giants and Tennessee
Titans fan, so these are badtimes for me in pro football.
Hey, tom, you have beentraveling the country, if not
(04:54):
the globe, going to variousconferences with your ear on the
ground.
What?
Tom Goldsby (05:01):
are some of the things you're hearing about in
supply chain these days.
Thanks, ted, and that's why Imissed the ad board, which, by
the way, I heard that was thebest ad board ever by some
estimation.
So I was really sorry to missthat.
But yeah, I've been out andabout.
Just got in late last night,found myself earlier this week
in Cameron Indoor Stadium of allplaces at Duke University, some
place I didn't know if I'd everbe, but that's neither here nor
(05:24):
there.
I've been out in some industrymeetings in the defense sector,
retail, also a National ScienceFoundation meeting on circular
economy last week at Vanderbilt,also a logistics doctoral
symposium at Auburn.
Those have all been in the lasttwo, two and a half weeks.
A lot of traveling.
It's really interesting.
I think back to the travels Ihad very early in the year.
(05:45):
Supply Chain Leaders in Actionconvened in the second week of
January and I think there werevery bullish expectations, a lot
of optimism and expectations ofgrowth.
And wait, wait can you hear thatthe air seeping out of the
balloon Not so bullish right now.
A lot of uncertainty, even somesuggestions of the R word
(06:07):
recession floating out there.
Consumer sentiments are downand I'm kind of picking up on
some of that in the last week orso.
I know we're going to betalking a little bit about
tariffs, but floating thesetariffs, implementing them one
day and pulling them back thenext and then threatening to
reintroduce them a month later,doesn't really instill
confidence and people willing topull the trigger and invest.
(06:29):
And so this kind of carrots andstick approach well, it's all
stick, it seems, frankly is kindof stressing us all out and
squeezing a lot of theenthusiasm, confidence and
exuberance that we saw earlierin the year.
It's just amazing how quicklythings have turned and frankly,
as in the defense sector, wouldyou think that would be one
(06:51):
budget that might not gettouched by Elon and Doge, and
now we're seeing cuts there arebeing proposed anyway.
Ted Stank (06:58):
Thomas mentioned Marianne Wanamaker, who's going
to be a keynote speaker at theforum in April.
She also came and joined us atthe advisory board and she's got
a lot of inside perspective, ifyou will, given her previous
positions.
She said one thing that theTrump administration is really
always keeping their eye on isthe stock market.
You know, I think, that as oftoday, president Trump kind of
(07:21):
gave a moratorium to tariffs onautomotive industry and also
just a little while agoannounced that there was a one
month moratorium on the 25percent tariff on Mexico.
So I think one thing I wouldcaution everybody is not to jump
into the realm of mass hysteria.
I think what the stock marketis is a one particular dimension
of mass hysteria and everybodyreacts and you don't kind of
(07:44):
want to be running with thecrowd.
So I think my recommendation ona lot of this there's a lot of
question about whether we'regoing to invest.
Jobs reports comes out tomorrowand there's a question as to
whether it's going to show thatwe added 30,000 jobs or 300,000
jobs.
Most of the federal layoffs willnot shown up yet in this jobs
report, so my recommendation issteady as she goes let's not get
(08:07):
into mass hysteria.
Tom Goldsby (08:09):
Yeah, I haven't cashed out the 401k, even though
there's been some suggestion ofit.
But you know, just in terms ofinvesting I do agree no need to
run toward a ledge or anythinglike that but just in terms of
the bullish outlook of just afew months ago, again, I think
we'll agree that floatingtariffs, pulling them back and
then threatening to introducethem time and again Again, it's
(08:31):
an easy thing to say.
Whether they actually collecteda dollar of tariffs is yet to
be seen, because they actuallyhadn't implemented them.
It takes a little time toactually implement policies.
I hear what you're saying, butagain reporting on the word on
the street.
For instance, I was supposed togo to Oklahoma City next week
for another military meeting.
(08:52):
Dod meeting canceled becausemilitary personnel are not
permitted to travel, and that'strue of a lot of federal folks.
They weren't at the NSF meetinglast week in Nashville, so
there's already starting to besome ramifications, I guess.
Ted Stank (09:04):
Sure, I totally agree , and I mean there's just
uncertainty across the boardwhether it's related in the
supply chain world.
Probably the thing that impactsus the most is what is
happening with tariffs.
I certainly wouldn't want to bein the automotive industry when
you think about how linked weare to Canada and Mexico, but
again, I heard a great commentfrom somebody who was
interviewed at the TPMconference that recently
(09:25):
convened in San Diego, and theysaid if you want to start
putting contingency plans inplace based on what happened
today, you'd be an absolute fool, because that may totally
change tomorrow.
You know, we said it duringCOVID Seems like we say it
during every crisis.
Resilience is the word of theday for supply chain managers,
and with this focus onresilience, we have tools
(09:47):
available to us today that wehaven't had in the past that I
think can help us, and so, withthat, we've asked Jason to join
us to talk about what I conceiveas the cutting edge of AI
applications in supply chain.
So, jason, welcome aboard Anyinitial comments you want to
make about what we've just beentalking about?
Jason Murray (10:05):
Well, great to be here.
Seriously, yeah, I mean, Ithink, largely speaking, supply
chains.
You either kind of get good atpredicting stuff, or if there's
a lot of uncertainty, you haveto be able to react fast, and I
think that's what we're seeing.
It's just how do you kind ofadjust to that.
I do tend to believe with you.
I'm kind of in the camp of.
(10:26):
I think Trump thinks that thestock market is very important,
and so it makes me veryskeptical of how far he's
willing to go, kind of from aneconomy perspective, regardless
of what your political bent is.
So that's my take on it.
I'm not ready to throw the yearin the garbage yet.
I think there's a lot thatneeds to play out first before
we can make any definitiveconclusion on where things are
going.
Ted Stank (10:45):
So I can't sell you my gold or my crypto?
No, yeah.
Jason Murray (10:49):
Yeah, I mean being at the forefront of technology.
I still don't totallyunderstand the crypto thing.
Ted Stank (10:57):
I'm glad to hear you say that.
Jason Murray (10:59):
I really have tried.
You know, the Bitcoin stuff isactually super interesting the
origination of that but the coinstuff is still just wild to me.
I don't understand it.
Thomas Deakins (11:08):
You don't have a bunch of meme coins, jason,
come on.
Jason Murray (11:11):
I don't.
Clearly I'm missing out, but Idon't, as it turns out.
Ted Stank (11:15):
So I'm glad to hear you say that, because I thought
I had entered the realm ofremember.
When you went to yourgrandparents' house, when we had
VCRs and the clock was flashing12 because they didn't have a
set, I thought that I wasgetting to that point.
Jason Murray (11:27):
Well, maybe, but I'm there with you if that's the
case.
Thomas Deakins (11:31):
Love it.
Let me chime in on one thingreal quick, because I saw this
this morning and I think this isprobably where we are with the
tariffs.
Brown Foreman was more worriedabout the fact that Canada was
going to pull Jack Daniels offthe shelf than they were about a
tariff.
That would hurt them more thanit would if a tariff got
implemented and I was just like,okay, there, we are right there
(11:51):
.
Yeah that's an absolute right.
It's the whiskey and thebourbon Canadians are going to
revolt yeah.
Ted Stank (11:56):
Thomas, you've known Jason for a long time.
You want to jump in and startasking some good questions of
Jason.
Thomas Deakins (12:01):
Yeah, let me just do real quick intro of
Jason for our listeners.
Jason's co-founder and CEO ofShipium, where he guides the
company's vision towardsbecoming the world's best supply
chain technology platform fore-commerce and retail.
Prior to founding Shipium,jason spent 19 years at Amazon
as VP of Retail Systems and VPof Forecasting and Supply Chain.
(12:21):
While there, he owned theglobal software and operations
group that powered Amazon Prime,subscribe and save and pricing.
He's a University of Washingtongrad and an engineer at heart,
who loves solving complex,scaling problems.
Jason Murray (12:35):
That about sums it up.
The high level story withShippium is I spent the 20 years
at Amazon 19 years and thewhole notion of using technology
to solve these types ofproblems has kind of become my
career, and Shippium is anextension of that.
You know I love talking abouteverything related to the space,
data science and now AI, andyou know before that maybe ML.
This whole trend I think isvery, very applicable to supply
(13:00):
chain, because the supply chainis all about prediction and
flexibility, Right, so it justfits in perfectly, and so all of
these discussions are kind ofmore about this larger trend of
what I see, what I've beenwatching happening.
I was lucky enough to watch ithappen at Amazon, but I'm
watching it continue after that.
I think, to some degree, we'rereally in a great time to be in
(13:21):
this space, because I think thetechnology is going to be
transformative and it just kindof keeps getting better and more
applicable.
Thomas Deakins (13:28):
What made you, you, or what was that trigger or
the catalyst that said, hey,I'm going to go co-found this
company called Shippium.
I guess what were the painpoints that you were seeing in
the industry?
Jason Murray (13:37):
You know, part of it is, I think, when I got out
of school I wanted to.
You know, I've always had kindof an interest in doing
something entrepreneurial and Iworked at a startup maybe a year
or two right out of schoolbefore Amazon and then got
recruited in Amazon.
At that point it was close to astartup.
(13:57):
It was very, very rough inthose early days in terms of
what it was and how big it wasand kind of where we fit in with
everything.
Then it obviously just kind ofblew up over the next several
years and you get caught up inthat and to some degree, me
leaving Amazon was largely aboutthis notion of I just kind of
wanted to get back to building,where you have a much larger
impact on the business, etcetera.
And when I left Amazon I wasn'ta hundred percent sure what I
wanted to do next.
I had a couple ideas butultimately when I started I did
(14:20):
some consulting type gigs andwhen I looked around the
industry, I think what you sawwas just that Amazon had built
this tool set, this technologystack that was so powerful and
useful if you're trying to run asupply chain and the way they
thought about stuff.
This was like a 20-year journey, right.
Kind of the approach, with theuse of data being a big point,
(14:40):
some of the other points beingthe coordination between these
stages and kind of more of aunified approach, et cetera was
just so powerful.
And then you kind of went andlooked at other retailers that's
who I was mostly talking to atthe time through my contacts,
and it was just, you know, thecomparison was like I always
kind of use the planet of theapes analogy, where you know
Amazon's got these rocket shipsand then you've got on the
(15:02):
ground people are fighting withsticks and stones, right.
And so the concept behindChipium really just started, a
form of if we can use some sortof democratization type
methodology where we're pullingtogether from several folks and
combining technology, combiningdata, we could build something
that kind of makes theseretailers competitive.
Since then it's expanded a bit,had a lot of traction with 3PLs
(15:25):
, for example, and we're nowkind of getting deeper into the
B2B space, which to some degreeI always kind of refer to as my
opinion, going through a littlebit of an Amazonification of
that area.
Supply chains are getting morecomplicated and they need to
adjust to that also, but that'sthe high level story.
That's how I kind of got here.
Thomas Deakins (15:41):
Ted, you might know somebody that used to work
at Amazon too, right?
Ted Stank (15:44):
Yeah, jason, you know one of our close friends is
Dave Clark.
He's been a guest on thispodcast before and,
interestingly, Dave is startingto make some waves in a similar
space with his new startupcalled Augur, where he's trying
to really create a platform thatpulls together.
I've been doing this 35 yearsand I've been talking about this
for 35 years and maybe we'regoing to see reality with what
(16:05):
Dave's doing and with whatyou're doing, which is the
ability to pull together datafrom across the supply chain so
that we really can I like whatyou said democratize things and
be able to create economies ofscale, if you will, across
different shippers, just becauseof being able to have that data
availability.
Jason Murray (16:23):
I think we need as many eyes on the space as
possible.
I worked for Dave for a coupleyears back in probably 2015 or
something that'd be my guess butI think the idea makes a ton of
sense, especially with thetechnology the latest iterations
of AI being really good atprocessing and transforming data
in such a way that you couldspeed this process up.
(16:44):
It sounds like they're stillkind of trying to figure out
exactly what it looks like, butthe concept, like the North Star
, makes a lot of sense to me.
Thomas Deakins (16:52):
Speaking of lots of data and being able to use
artificial intelligence in thatsetting.
I did some research and foundout that in 2024, jason correct
me if I'm wrong the Shippiumplatform you guys made over a
billion that's a billion with aB predictions in the platform.
You guys made over a billionthat's a billion with a B
(17:13):
predictions in the platform, andI know that's across numerous
industries like retail,automotive, healthcare and, as
you said earlier, a lot of yourclients are in the 3PL sector,
or we'd call it the LSB sector,but talk a little bit about that
, because what's that mean fromwhen I use the word prediction?
So, for our listeners, talk alittle bit about the shipping
platform.
What are you doing for yourcustomers?
(17:33):
What's the main problem you'retrying to solve?
Jason Murray (17:36):
If I kind of like back out a little bit and just
talk about the approach.
I think the approach is not newper se, right, this is just
coming directly from kind of theAmazon playbook.
The idea here is you keepfinding new ways to get as much
data as possible and then youuse that data to help make
decisions down through thiswhole pipeline At a high level.
(17:57):
That's simply kind of whatwe're doing and we've had the
most luck with the company'sbeen through iterations, as you
do in early stage companies, andwe thought that the way into
people was going to be kind ofmore on the customer experience
side.
What we found when we got onthe ground was there's so many
to be kind of more on thecustomer experience side.
What we found when we got onthe ground was there's so many
problems in kind of just thebasic shipping optimization
(18:17):
piece that that's really wherethey wanted to help, and so
that's where we've had kind ofour biggest success right, and
so we help with a lot ofdecisions around things like
transportation, laneoptimization, where to ship from
, trying to hit target dateswith your shipping methods.
As this data has expanded inour platform, we're now kind of
predicting transit times of allthese carriers and methods.
(18:39):
So we can do a really finegrained job of either telling
the customer this is when youshould expect it or knowing that
if I use this given method orlane, it's going to hit within
this timeframe.
And then the other part that'sbeen really interesting is just
kind of you take all that datathat you've acquired and then
allow people to run scenarioswith it, right, and then have
the tooling to do that, and wecall that product on our side
(19:01):
simulation.
But the point just to kind ofback it up right again, the
magic happens with cloudtechnology enabled and then
storage getting cheaper and nowthe massive spend and expansion
of processing units for workingthrough these data in more
predictive ways has enabled isyou can just start pulling more
and more data into theseplatforms and then use that to
(19:21):
make better decisions.
I kind of emphasize that in asimple way.
I think that is something thatthis industry just has not been
able to get started on.
It sounds like painfully simpleand it's kind of the standard
playbook within the techindustry.
But there's something about thefact that if you kind of came
up through even the techplatforms that grew up in the
(19:43):
80s and 90s and they've just hadtrouble converting to that
method of thought.
They're all trying To somedegree.
I think of it as almost thisDNA thing, where you have to
kind of build the platform withthat in mind from the start and
then you get better outcomes.
And that's why I think it'skind of exciting to see this new
crop of technology platforms inthis space and kind of seeing
what the approach is right,whether that be Flexport, which
(20:11):
is kind of doing a more of atech-enabled model to us, which
we are just like laser, focusedon just the tech and data and
optimization.
Ted Stank (20:14):
So, jason, when you came and spoke to the advisory
board, you did a talk about howdo you wade into this wild west
world of technology andspecifically generative AI,
agentive AI.
Can you give us a thumbnailsketch of your perspective on
that, how you all are using itand if somebody is out there
listening and saying, yeah, weknow that we need to get going
on it, maybe some advice on howto jump in and move with it?
Jason Murray (20:37):
Yeah, I think a big part of it is like cultural.
I mean that in the sense thatthis is kind of part of the
issue.
Chances are, whatever stateyou're in now, things are going
to get a little worse beforethey get better as you go into
this.
So that takes a certain levelof kind of boldness as a leader
to say like we're going tocommit to solving this problem
and actually put resources to it.
It's explored in things likeInnovator's Dilemma and there's
(21:00):
multiple business books on thistopic.
But I think if you have thisapproach, if your culture of
your company is built around I'mgoing to make decisions on gut
feeling and history and kind ofwhat we've done in the past and
it's worked, et cetera, you'regoing to kind of get that as an
outcome.
One of the more powerful aspectsof Amazon was the willingness
to kind of actually follow thedata.
It's never perfect.
(21:21):
Humans have biases and thisisn't easy and it takes really
smart people to determine whensomething is correlated versus
causal, et cetera, et cetera.
You can go on and on.
But if there's a commitmentculturally from the company to
kind of lean into, we want tomove towards automation,
optimization, use of data tomake decisions, you will start
to see a shift.
That's kind of number one.
(21:42):
And then I think you have thesekind of tactical follow-ups,
which are interesting, aroundinvesting in storage of data and
capturing of data, working withdata forward vendors, you know,
that's kind of how we fit in.
But all of that kind of anchorsback to this first piece that
you've got to start with acommitment.
You have to believe this is theway to move the company forward
.
Ted Stank (22:02):
Correct me if I'm wrong, but my experience is that
once you start wading in andgetting kind of up to your waist
, your only limit is your team'sinnovativeness and willingness
to try certain things.
You know the whole cliche offail fast, yeah, but you know
just to say, hey, can we usethis for this and to speed this
(22:22):
up and to make us more accurate.
Is that kind of a right way tolook at it?
Jason Murray (22:27):
Absolutely.
It starts building on itself,right.
You capture one piece of dataand that leads to another.
That leads to how you're makingdecisions here and this is
going to help with thesedecisions here, and it builds on
itself, right, and that'sthat's kind of why I talk about
culture and and I think that'sthe only way to do it you can
definitely overcook it on theplanning side.
People just tend to overthink.
It's not like technologists areimmune to this.
(22:50):
I think two years ago, whenchat GPT blew up, there was 500
predictions of where it wasgoing to go, and most of those
haven't actually happened.
It's gone all sorts of otherdirections, right?
I think it's very hard topredict.
You have to kind of be activeand trying to play with it, but
you wait in, even if the thingyou're doing isn't working.
Specifically, you're nowlearning how to operate in this
(23:12):
environment.
Which is going to capture thenext thing that comes online, et
cetera.
Tom Goldsby (23:16):
Hey Jason, I'm going to be serving on a panel
at the ProMat in Chicago inabout 10 days, looking forward
to that, alongside John Paxson,the CEO of MHI, and they gave me
an advanced copy of the annualreport that MHI generates each
year and they said Gen AI isgreat.
We need to figure out how toleverage it.
But the next wave is agentic AI.
(23:39):
Can you help set the table alittle bit, maybe prepare me for
this panel so I can soundintelligent for that session?
Jason Murray (23:44):
coming up, I think , of agentic AI.
It's a bit of a fuzzydefinition, so I'll just give
you how I kind of think about it.
If you were to replace some jobin your company with an AI bot
or an AI agent.
That's how I kind of thinkabout agentic, right, it's a
little disingenuous in the sensethat maybe an example would be
a customer service rep who canrespond to things.
(24:05):
Another example would besomeone who's kind of constantly
going out and testing thesystem autonomously.
Another one would be someonewho's kind of constantly going
out and testing the systemautonomously.
Another one would be likeextracting data from.
You know, I'm going to go tothese bunch of carriers'
websites and try to extract datathat might be useful in
decisioning.
Or imagine like a person usedto have to do that manually and
this thing is just going andkind of summarizing and pulling
(24:26):
it together and bringing it intothe fray, right.
Those are examples of agenticAI.
I think the concept ofautonomous demons or things
running in software that wentout and did stuff that's
obviously been around.
I think what makes it unique iskind of the notion that they're
able to kind of self-adjust alittle bit more than they have
in the past.
That's what makes itinteresting If I'm getting
customer service requests fromsomeone, I'm actually going to
(24:48):
adjust, you know, try to processthrough and make kind of
adjustments to how I'mresponding in a more autonomous
way than like we have been ableto in the past.
And that's a function of justthe power of these models that
are coming online and theability to kind of reason
through some of that stuff.
Ted Stank (25:04):
Hey, Tom, just to be clear, you're going to be on a
panel with John Paxton, CEO ofMHI, right?
Not John Paxson, formerteammate of Michael Jordan on
the Bulls, who used to swishthose threes all the time.
Tom Goldsby (25:18):
That's helpful.
You know, I think just gettingmy pronunciation of agentic AI
is going to be a starting pointfor me, but that's a great
explanation.
And do you see thosecapabilities making their way
into Shippium and what you haveto offer?
Jason Murray (25:31):
You know, I would say this new wave of AI is
largely about automation.
There's so much pressure aroundreplacing jobs and there's so
much energy around that right,and a lot of it's able to
automate these basic tasks thatpeople had to do before.
If I go back to my originalcomment on just the whole thing
is we're just trying to pull inas much data as possible.
The challenge you have insupply chain currently is, you
(25:54):
know, there's like several tensof thousands of carriers in the
United States alone and somemassive number past that and on
the world right, and so they alldon't have EDI or API access to
them.
The thought of can I go out andextract data from these and
pull that into this decisioningprocess, that's been super
interesting.
You know.
Obviously this was maybe moreon the straddling the line
(26:16):
between what you'd callgenerative and agentic, but the
coding stuff is prettyincredible, right.
I mean that's been both fromthe standpoint of our developer
productivity, how much we'reable to get done with a given
developer.
On top of that, the thinkingabout part of the challenge you
have is just integrating.
When I started Chipium, Ithought we build a product and
then people use it right.
Well, if you're really in thespace.
(26:38):
A huge part of the problem isintegrating with people, and so
we've had a lot of luck withbeing able to simplify that
process, and at least the AI isable to help do this initial
mapping, which speeds up thisprocess.
So I think the more this stuffcomes, you're just going to see
more and more use cases as itcontinues to develop and kind of
(26:58):
build out.
You know we have a very kind ofopen approach to it, like I
will try anything.
Sometimes it doesn't work, andthen you just perhaps wait a
little while until you hearsomething else.
But we try to be very open toeverything, and I think there's
a lot of use cases are appearingas we keep digging into this.
Thomas Deakins (27:14):
Good deal.
Let me ask you one morequestion here real quick.
Look into your crystal ball,jason, and I want you to tell me
, like where is the Shippiumroadmap, based on what you're
hearing from the industry andtalking to your customers and I
know you're heavily involved onthe prospect side and sales
cycles too.
But what's the next evolutionof?
Jason Murray (27:33):
Shippium, we've done a really good job on kind
of what I would call theshipping optimization space.
It centers around understandingthat I think we're planning on
pushing in both directions right.
We just kind of launched ourinitial billing management
product for 3PLs, where nowwe're getting into kind of more
some of the invoicing side andhow they mark stuff up, and this
(27:53):
is actually another example ofa great use of Gentic AI, in the
sense that we can potentiallyhave something that is looking
for anomalies and speed up thatinvoicing process in a much more
automated way than it's done inthe past.
You go the other direction, Ithink.
You kind of head more towardsthe customer experience side,
and this goes back to like wherewe started.
(28:13):
The idea is is that we continueto push up into delivery
promise and driving growththrough the platform, as opposed
to being purely about costoptimization.
And then I think the third thingI kind of mentioned, we've
gotten into simulation and havethese large systems that can
(28:41):
just use the same executionengine to process virtual
transactions to come up with,potentially.
What does this lead to from aplanning perspective?
And so we see ourselves aspushing on all fronts.
In a lot of cases it will meanpartnering with people and
perhaps like, if they're good atsome aspect of it, we're able
to kind of take maybe thetransportation piece and push
that into their world also.
But all systems go in terms of.
There just seems to be a lot ofrunway in all directions.
Tom Goldsby (29:03):
That's awesome.
You know, and Ted, this had tocross your mind as Jason was
describing the value propositionof applying AI to complex
transportation problems.
Remember that paper we wroteway back in the year 2000, 25
years ago, yeah, and it was justa concept of all the cool
things we were hoping to do.
Is the future now?
(29:24):
Are we living in that time?
Ted Stank (29:26):
You know, I think like everything, tom, it's
evolving.
Ten years from now there'llprobably be other technology
that can come to play, but youknow, the concepts of supply
chain management haven't changedtremendously.
You know, in a couple ofmillennium probably, but the
technology we apply to make itmore efficient, more effective,
certainly has and will continueto.
So I think one of the reallycool things that I've seen in my
(29:52):
career is the number of peoplewith Jason's kind of background,
this solid computer engineeringbackground, coming into the
supply chain world.
Like I'll have students come upto me and say I'm a supply
chain major but I'm thinkingabout doing business analytics
as a minor.
Do you think that's a goodthing?
I'm like, dude, you want to geta good six figure opening
salary coming out of theundergrad program and, yeah, do
that.
That's a really good thing.
(30:14):
Not to pick on my marketingcolleagues, but so many of our
students for a long time weresupply chain majors, marketing
minors, and I always tell themwell, you have a good career if
you want to go into sales with a3PL, but if you really want to
roll up your sleeves and getinvolved with where the world is
going, it's really with thedata side, the other thing I'd
like to highlight.
Jason, you mentioned culture.
(30:34):
So you're not just pureengineering anymore.
You've learned right.
I have my roots in engineering,my roots in operations, and you
know, I used to just thinkculture was this bogus stuff
that psychologists talked about.
But man, culture is everything100%.
Jason Murray (30:49):
As Drucker said, culture eats strategy.
Right, a lot, you know, andwe're a long ways away from
people not being involved.
You need the motivation and thepush and all the other stuff,
and that's a critical part of anorganization is anything.
You can have all the technologyyou want, but if it's not being
used or embraced in the rightway, it won't go anywhere.
You see that over and over.
Ted Stank (31:10):
That's why it's all about leadership, right.
Jason Murray (31:12):
Yeah, exactly.
Tom Goldsby (31:14):
Well, I think we got to wrap up about now.
But hey, Jason, thank you somuch.
We Well, I think we got to wrapup about now, but hey, Jason,
thank you so much.
We've been.
You know this has been a datethat's been circled on our
calendar for some time andthings are changing so rapidly.
So thank you very much forsorting things out for us and
giving a sense for where we are,the potential of AI and also
where it's heading.
And hey, Thomas, thank you verymuch for joining us.
As well as a co-host, you wereinstrumental in bringing Jason
(31:38):
into our fold.
All as a co-host, you wereinstrumental in bringing Jason
into our fold, so we reallyappreciate you for that and
again, elevating our level ofplay, just as our team around
GSCI does each and every day.
And hey, just to give a littlebit of a teaser of what's in
store, the next supply chainforum is coming up in early
April and, Ted, I think we'vegot a special guest.
Colin Yankee, the Chief SupplyChain Officer of Tractor Supply,
(32:00):
is going to be our specialguest on the next podcast.
Right?
Ted Stank (32:03):
Absolutely.
We'll be doing our next podcastlive from the forum on April
9th, I think.
Is that right, Thomas?
I think so.
April 9th with Colin Yankeefrom Tractor Supply.
Tom Goldsby (32:13):
Another great partner of ours, Jason.
Ted Stank (32:14):
I echo my thanks to you and to Thomas.
Great to have you guys with us.
Wish we were all togetherbecause I would go out and buy a
round for us all.
Jason Murray (32:23):
Oh it sounds good.
Ted Stank (32:24):
Maybe you can do it and put in a request on Venmo
for me to pay you for it.
Tom Goldsby (32:29):
Yes.
No pappy though no pappy Topshelf, jason, you earned it, man
, all right everybody.
Ted Stank (32:35):
Thanks, as always.
If anybody has any questions,suggestions, you can reach us at
gsciutkedu Until next time bewell.
Intro & Outro (32:44):
Thanks for tuning in to Tennessee on Supply Chain
Management.
If you enjoyed the episode,subscribe today on your favorite
listening platform to get allof our episodes as soon as they
drop, and don't forget to take amoment to leave us a rating.
Have any questions, thoughts orfeedback?
We'd love to hear from ourlisteners.
Email us at gsci at utkedu.
Join us next time as wecontinue pulling back the
(33:07):
curtain on the world of supplychain, educating and
entertaining you along the way.
Until then, listeners.
Welcome to the Tennessee on Supply Chain
Management podcast.
Listen in as co-hosts Ted Stankand Tom Goldsby set sail into
the world of end-to-end supplychain management, diving deep
into today's most relevantbusiness topics.
They'll share insights inpressing industry issues and
tackle the challenges keepingsupply chain professionals up at
night.
If you're enjoying the ride,download and subscribe to
(00:22):
Tennessee on Supply ChainManagement on your favorite
podcast platform now.
Ted Stank (00:28):
Hello everyone and welcome to Tennessee on Supply
Chain Management.
This is our sixth episode ofseason three.
A lot of crazy stuff going on.
As we sit here recording, it'searly March 2025, six weeks into
the new Trump 2.0administration, and if anybody
is not on the far side of themoon, you'll know that there is
(00:50):
just a lot of change going on.
Much of that change isimpacting the supply chain.
We'll touch on some of thattoday and then we'll talk about
how do we stay flexible?
What are tools we can use totry to help us?
In this time, we'll bring onour guest, jason Murray, to talk
about AI applications.
In this time, we'll bring onour guest, jason Murray, to talk
about AI applications in supplychain.
Before we get into that, let'sintroduce my co-host, great
(01:11):
friend, tom Goldsby.
Tom Goldsby (01:12):
Hello, ted, great to be back with you.
Our loyal listeners may haverealized that I was missing on
the last episode, butfortunately we had Josh King,
the Director of Marketing andOperations for the Global Supply
Chain Institute here at UT, dosuch a great job on the last
podcast.
That said, I'm glad to be backin the mix, but hey, we've got a
(01:34):
real full room today, so let'sget around to introducing some
of the other folks.
Ted Stank (01:39):
Good deal.
One of the other people joiningus as co-host is our own Thomas
Deakins.
Thomas is Managing Director ofour own Thomas Deakins Thomas is
managing director of our GlobalSupply Chain Institute.
He's also a UT supply chainmanagement undergrad and MBA
former vice president of Project44, and also worked with
companies like Oracle, G-Log,Trimble, Mavenwire, MercuryGate,
(02:00):
CHEP, SLI Consulting and JBHunt, so pretty much covered the
spectrum of supply chaincompanies, Thomas, welcome.
Thomas Deakins (02:08):
Hey, happy to be here and honored to be part of
the show, so looking forward toit.
Ted Stank (02:14):
Great Also like to welcome our guest, Jason Murray.
Jason is co-founder and CEO atShippium after a 19-year career
and ended up as a vice presidentwith Amazon.
Jason's been with us as apartner on Global Supply Chain
Institute for a while now.
He spoke to our executiveadvisory board last year and
just blew us all away with whathe's doing at Chipium and the
(02:35):
promise of AI in helping us inuncertain times.
Again, as I hedged that early,if you don't think that we're
living in uncertain times, thenyou probably have been on the
far side of the moon, Jason,welcome.
Jason Murray (02:49):
Hey, great to be here.
Nice to see you guys, I'mexcited.
Ted Stank (02:51):
Hey, I mentioned a little while ago that we had our
executive advisory board.
We recorded our podcast fromthat executive advisory board
last month.
A lot of great discussion.
One of the big topics was howdo we apply AI in the supply
chain?
I think it's one of thoseissues that we all know it's
coming.
Everybody is trying it indifferent ways.
I'm not sure yet that there aresome great use cases, although
(03:14):
hopefully Jason will be able toshed some light on that.
We had about 30 of our advisoryboard members present, had some
great presentations on some ofthe projects we're working on in
the Advanced Supply Chain,collaborative around planning,
around how do we improveretention of Gen Y and Gen Z
labor, particularlymanufacturing and operations.
(03:35):
So a lot of really good stuff.
It was great to have them allthere.
Thomas, we also have somethingbig coming up in the second week
in April.
You want to talk a little bitabout Forum.
Thomas Deakins (03:44):
Yeah, thanks, ted.
April 8th through 10th we willhave our Supply Chain Forum.
It'll be the one for the spring.
Downtown Knoxville Marriottagain dates April 8th through
the 10th.
So if you haven't registered,please do so.
Look forward to seeingeverybody.
We've got a great lineup ofcontent.
Two of the guest speakers I'llhighlight real quick.
(04:06):
One will be Marianne Wanamaker,who's the Dean of the Baker
School for Policy and formereconomist at the White House in
the Trump 1.0.
And she will give us a state ofthe economy, so we're so
excited to hear about that.
And then the second one is thenew voice of the Vols, mike
Keith.
Mike used to be the voice ofthe Titans.
(04:28):
It'll be great to hear from himand also just the ties that he
has back to our department.
Ted Stank (04:33):
I think Mike came over to Tennessee because the
Vols are looking a lot betterthan the Titans these days.
That's not hard.
The Titans used to playfootball, but I'm not sure
anymore.
This is said as a person who isa New York Giants and Tennessee
Titans fan, so these are badtimes for me in pro football.
Hey, tom, you have beentraveling the country, if not
(04:54):
the globe, going to variousconferences with your ear on the
ground.
What?
Tom Goldsby (05:01):
are some of the things you're hearing about in
supply chain these days.
Thanks, ted, and that's why Imissed the ad board, which, by
the way, I heard that was thebest ad board ever by some
estimation.
So I was really sorry to missthat.
But yeah, I've been out andabout.
Just got in late last night,found myself earlier this week
in Cameron Indoor Stadium of allplaces at Duke University, some
place I didn't know if I'd everbe, but that's neither here nor
(05:24):
there.
I've been out in some industrymeetings in the defense sector,
retail, also a National ScienceFoundation meeting on circular
economy last week at Vanderbilt,also a logistics doctoral
symposium at Auburn.
Those have all been in the lasttwo, two and a half weeks.
A lot of traveling.
It's really interesting.
I think back to the travels Ihad very early in the year.
(05:45):
Supply Chain Leaders in Actionconvened in the second week of
January and I think there werevery bullish expectations, a lot
of optimism and expectations ofgrowth.
And wait, wait can you hear thatthe air seeping out of the
balloon Not so bullish right now.
A lot of uncertainty, even somesuggestions of the R word
(06:07):
recession floating out there.
Consumer sentiments are downand I'm kind of picking up on
some of that in the last week orso.
I know we're going to betalking a little bit about
tariffs, but floating thesetariffs, implementing them one
day and pulling them back thenext and then threatening to
reintroduce them a month later,doesn't really instill
confidence and people willing topull the trigger and invest.
(06:29):
And so this kind of carrots andstick approach well, it's all
stick, it seems, frankly is kindof stressing us all out and
squeezing a lot of theenthusiasm, confidence and
exuberance that we saw earlierin the year.
It's just amazing how quicklythings have turned and frankly,
as in the defense sector, wouldyou think that would be one
(06:51):
budget that might not gettouched by Elon and Doge, and
now we're seeing cuts there arebeing proposed anyway.
Ted Stank (06:58):
Thomas mentioned Marianne Wanamaker, who's going
to be a keynote speaker at theforum in April.
She also came and joined us atthe advisory board and she's got
a lot of inside perspective, ifyou will, given her previous
positions.
She said one thing that theTrump administration is really
always keeping their eye on isthe stock market.
You know, I think, that as oftoday, president Trump kind of
(07:21):
gave a moratorium to tariffs onautomotive industry and also
just a little while agoannounced that there was a one
month moratorium on the 25percent tariff on Mexico.
So I think one thing I wouldcaution everybody is not to jump
into the realm of mass hysteria.
I think what the stock marketis is a one particular dimension
of mass hysteria and everybodyreacts and you don't kind of
(07:44):
want to be running with thecrowd.
So I think my recommendation ona lot of this there's a lot of
question about whether we'regoing to invest.
Jobs reports comes out tomorrowand there's a question as to
whether it's going to show thatwe added 30,000 jobs or 300,000
jobs.
Most of the federal layoffs willnot shown up yet in this jobs
report, so my recommendation issteady as she goes let's not get
(08:07):
into mass hysteria.
Tom Goldsby (08:09):
Yeah, I haven't cashed out the 401k, even though
there's been some suggestion ofit.
But you know, just in terms ofinvesting I do agree no need to
run toward a ledge or anythinglike that but just in terms of
the bullish outlook of just afew months ago, again, I think
we'll agree that floatingtariffs, pulling them back and
then threatening to introducethem time and again Again, it's
(08:31):
an easy thing to say.
Whether they actually collecteda dollar of tariffs is yet to
be seen, because they actuallyhadn't implemented them.
It takes a little time toactually implement policies.
I hear what you're saying, butagain reporting on the word on
the street.
For instance, I was supposed togo to Oklahoma City next week
for another military meeting.
(08:52):
Dod meeting canceled becausemilitary personnel are not
permitted to travel, and that'strue of a lot of federal folks.
They weren't at the NSF meetinglast week in Nashville, so
there's already starting to besome ramifications, I guess.
Ted Stank (09:04):
Sure, I totally agree , and I mean there's just
uncertainty across the boardwhether it's related in the
supply chain world.
Probably the thing that impactsus the most is what is
happening with tariffs.
I certainly wouldn't want to bein the automotive industry when
you think about how linked weare to Canada and Mexico, but
again, I heard a great commentfrom somebody who was
interviewed at the TPMconference that recently
(09:25):
convened in San Diego, and theysaid if you want to start
putting contingency plans inplace based on what happened
today, you'd be an absolute fool, because that may totally
change tomorrow.
You know, we said it duringCOVID Seems like we say it
during every crisis.
Resilience is the word of theday for supply chain managers,
and with this focus onresilience, we have tools
(09:47):
available to us today that wehaven't had in the past that I
think can help us, and so, withthat, we've asked Jason to join
us to talk about what I conceiveas the cutting edge of AI
applications in supply chain.
So, jason, welcome aboard Anyinitial comments you want to
make about what we've just beentalking about?
Jason Murray (10:05):
Well, great to be here.
Seriously, yeah, I mean, Ithink, largely speaking, supply
chains.
You either kind of get good atpredicting stuff, or if there's
a lot of uncertainty, you haveto be able to react fast, and I
think that's what we're seeing.
It's just how do you kind ofadjust to that.
I do tend to believe with you.
I'm kind of in the camp of.
(10:26):
I think Trump thinks that thestock market is very important,
and so it makes me veryskeptical of how far he's
willing to go, kind of from aneconomy perspective, regardless
of what your political bent is.
So that's my take on it.
I'm not ready to throw the yearin the garbage yet.
I think there's a lot thatneeds to play out first before
we can make any definitiveconclusion on where things are
going.
Ted Stank (10:45):
So I can't sell you my gold or my crypto?
No, yeah.
Jason Murray (10:49):
Yeah, I mean being at the forefront of technology.
I still don't totallyunderstand the crypto thing.
Ted Stank (10:57):
I'm glad to hear you say that.
Jason Murray (10:59):
I really have tried.
You know, the Bitcoin stuff isactually super interesting the
origination of that but the coinstuff is still just wild to me.
I don't understand it.
Thomas Deakins (11:08):
You don't have a bunch of meme coins, jason,
come on.
Jason Murray (11:11):
I don't.
Clearly I'm missing out, but Idon't, as it turns out.
Ted Stank (11:15):
So I'm glad to hear you say that, because I thought
I had entered the realm ofremember.
When you went to yourgrandparents' house, when we had
VCRs and the clock was flashing12 because they didn't have a
set, I thought that I wasgetting to that point.
Jason Murray (11:27):
Well, maybe, but I'm there with you if that's the
case.
Thomas Deakins (11:31):
Love it.
Let me chime in on one thingreal quick, because I saw this
this morning and I think this isprobably where we are with the
tariffs.
Brown Foreman was more worriedabout the fact that Canada was
going to pull Jack Daniels offthe shelf than they were about a
tariff.
That would hurt them more thanit would if a tariff got
implemented and I was just like,okay, there, we are right there
(11:51):
.
Yeah that's an absolute right.
It's the whiskey and thebourbon Canadians are going to
revolt yeah.
Ted Stank (11:56):
Thomas, you've known Jason for a long time.
You want to jump in and startasking some good questions of
Jason.
Thomas Deakins (12:01):
Yeah, let me just do real quick intro of
Jason for our listeners.
Jason's co-founder and CEO ofShipium, where he guides the
company's vision towardsbecoming the world's best supply
chain technology platform fore-commerce and retail.
Prior to founding Shipium,jason spent 19 years at Amazon
as VP of Retail Systems and VPof Forecasting and Supply Chain.
(12:21):
While there, he owned theglobal software and operations
group that powered Amazon Prime,subscribe and save and pricing.
He's a University of Washingtongrad and an engineer at heart,
who loves solving complex,scaling problems.
Jason Murray (12:35):
That about sums it up.
The high level story withShippium is I spent the 20 years
at Amazon 19 years and thewhole notion of using technology
to solve these types ofproblems has kind of become my
career, and Shippium is anextension of that.
You know I love talking abouteverything related to the space,
data science and now AI, andyou know before that maybe ML.
This whole trend I think isvery, very applicable to supply
(13:00):
chain, because the supply chainis all about prediction and
flexibility, Right, so it justfits in perfectly, and so all of
these discussions are kind ofmore about this larger trend of
what I see, what I've beenwatching happening.
I was lucky enough to watch ithappen at Amazon, but I'm
watching it continue after that.
I think, to some degree, we'rereally in a great time to be in
(13:21):
this space, because I think thetechnology is going to be
transformative and it just kindof keeps getting better and more
applicable.
Thomas Deakins (13:28):
What made you, you, or what was that trigger or
the catalyst that said, hey,I'm going to go co-found this
company called Shippium.
I guess what were the painpoints that you were seeing in
the industry?
Jason Murray (13:37):
You know, part of it is, I think, when I got out
of school I wanted to.
You know, I've always had kindof an interest in doing
something entrepreneurial and Iworked at a startup maybe a year
or two right out of schoolbefore Amazon and then got
recruited in Amazon.
At that point it was close to astartup.
(13:57):
It was very, very rough inthose early days in terms of
what it was and how big it wasand kind of where we fit in with
everything.
Then it obviously just kind ofblew up over the next several
years and you get caught up inthat and to some degree, me
leaving Amazon was largely aboutthis notion of I just kind of
wanted to get back to building,where you have a much larger
impact on the business, etcetera.
And when I left Amazon I wasn'ta hundred percent sure what I
wanted to do next.
I had a couple ideas butultimately when I started I did
(14:20):
some consulting type gigs andwhen I looked around the
industry, I think what you sawwas just that Amazon had built
this tool set, this technologystack that was so powerful and
useful if you're trying to run asupply chain and the way they
thought about stuff.
This was like a 20-year journey, right.
Kind of the approach, with theuse of data being a big point,
(14:40):
some of the other points beingthe coordination between these
stages and kind of more of aunified approach, et cetera was
just so powerful.
And then you kind of went andlooked at other retailers that's
who I was mostly talking to atthe time through my contacts,
and it was just, you know, thecomparison was like I always
kind of use the planet of theapes analogy, where you know
Amazon's got these rocket shipsand then you've got on the
(15:02):
ground people are fighting withsticks and stones, right.
And so the concept behindChipium really just started, a
form of if we can use some sortof democratization type
methodology where we're pullingtogether from several folks and
combining technology, combiningdata, we could build something
that kind of makes theseretailers competitive.
Since then it's expanded a bit,had a lot of traction with 3PLs
(15:25):
, for example, and we're nowkind of getting deeper into the
B2B space, which to some degreeI always kind of refer to as my
opinion, going through a littlebit of an Amazonification of
that area.
Supply chains are getting morecomplicated and they need to
adjust to that also, but that'sthe high level story.
That's how I kind of got here.
Thomas Deakins (15:41):
Ted, you might know somebody that used to work
at Amazon too, right?
Ted Stank (15:44):
Yeah, jason, you know one of our close friends is
Dave Clark.
He's been a guest on thispodcast before and,
interestingly, Dave is startingto make some waves in a similar
space with his new startupcalled Augur, where he's trying
to really create a platform thatpulls together.
I've been doing this 35 yearsand I've been talking about this
for 35 years and maybe we'regoing to see reality with what
(16:05):
Dave's doing and with whatyou're doing, which is the
ability to pull together datafrom across the supply chain so
that we really can I like whatyou said democratize things and
be able to create economies ofscale, if you will, across
different shippers, just becauseof being able to have that data
availability.
Jason Murray (16:23):
I think we need as many eyes on the space as
possible.
I worked for Dave for a coupleyears back in probably 2015 or
something that'd be my guess butI think the idea makes a ton of
sense, especially with thetechnology the latest iterations
of AI being really good atprocessing and transforming data
in such a way that you couldspeed this process up.
(16:44):
It sounds like they're stillkind of trying to figure out
exactly what it looks like, butthe concept, like the North Star
, makes a lot of sense to me.
Thomas Deakins (16:52):
Speaking of lots of data and being able to use
artificial intelligence in thatsetting.
I did some research and foundout that in 2024, jason correct
me if I'm wrong the Shippiumplatform you guys made over a
billion that's a billion with aB predictions in the platform.
You guys made over a billionthat's a billion with a B
(17:13):
predictions in the platform, andI know that's across numerous
industries like retail,automotive, healthcare and, as
you said earlier, a lot of yourclients are in the 3PL sector,
or we'd call it the LSB sector,but talk a little bit about that
, because what's that mean fromwhen I use the word prediction?
So, for our listeners, talk alittle bit about the shipping
platform.
What are you doing for yourcustomers?
(17:33):
What's the main problem you'retrying to solve?
Jason Murray (17:36):
If I kind of like back out a little bit and just
talk about the approach.
I think the approach is not newper se, right, this is just
coming directly from kind of theAmazon playbook.
The idea here is you keepfinding new ways to get as much
data as possible and then youuse that data to help make
decisions down through thiswhole pipeline At a high level.
(17:57):
That's simply kind of whatwe're doing and we've had the
most luck with the company'sbeen through iterations, as you
do in early stage companies, andwe thought that the way into
people was going to be kind ofmore on the customer experience
side.
What we found when we got onthe ground was there's so many
to be kind of more on thecustomer experience side.
What we found when we got onthe ground was there's so many
problems in kind of just thebasic shipping optimization
(18:17):
piece that that's really wherethey wanted to help, and so
that's where we've had kind ofour biggest success right, and
so we help with a lot ofdecisions around things like
transportation, laneoptimization, where to ship from
, trying to hit target dateswith your shipping methods.
As this data has expanded inour platform, we're now kind of
predicting transit times of allthese carriers and methods.
(18:39):
So we can do a really finegrained job of either telling
the customer this is when youshould expect it or knowing that
if I use this given method orlane, it's going to hit within
this timeframe.
And then the other part that'sbeen really interesting is just
kind of you take all that datathat you've acquired and then
allow people to run scenarioswith it, right, and then have
the tooling to do that, and wecall that product on our side
(19:01):
simulation.
But the point just to kind ofback it up right again, the
magic happens with cloudtechnology enabled and then
storage getting cheaper and nowthe massive spend and expansion
of processing units for workingthrough these data in more
predictive ways has enabled isyou can just start pulling more
and more data into theseplatforms and then use that to
(19:21):
make better decisions.
I kind of emphasize that in asimple way.
I think that is something thatthis industry just has not been
able to get started on.
It sounds like painfully simpleand it's kind of the standard
playbook within the techindustry.
But there's something about thefact that if you kind of came
up through even the techplatforms that grew up in the
(19:43):
80s and 90s and they've just hadtrouble converting to that
method of thought.
They're all trying To somedegree.
I think of it as almost thisDNA thing, where you have to
kind of build the platform withthat in mind from the start and
then you get better outcomes.
And that's why I think it'skind of exciting to see this new
crop of technology platforms inthis space and kind of seeing
what the approach is right,whether that be Flexport, which
(20:11):
is kind of doing a more of atech-enabled model to us, which
we are just like laser, focusedon just the tech and data and
optimization.
Ted Stank (20:14):
So, jason, when you came and spoke to the advisory
board, you did a talk about howdo you wade into this wild west
world of technology andspecifically generative AI,
agentive AI.
Can you give us a thumbnailsketch of your perspective on
that, how you all are using itand if somebody is out there
listening and saying, yeah, weknow that we need to get going
on it, maybe some advice on howto jump in and move with it?
Jason Murray (20:37):
Yeah, I think a big part of it is like cultural.
I mean that in the sense thatthis is kind of part of the
issue.
Chances are, whatever stateyou're in now, things are going
to get a little worse beforethey get better as you go into
this.
So that takes a certain levelof kind of boldness as a leader
to say like we're going tocommit to solving this problem
and actually put resources to it.
It's explored in things likeInnovator's Dilemma and there's
(21:00):
multiple business books on thistopic.
But I think if you have thisapproach, if your culture of
your company is built around I'mgoing to make decisions on gut
feeling and history and kind ofwhat we've done in the past and
it's worked, et cetera, you'regoing to kind of get that as an
outcome.
One of the more powerful aspectsof Amazon was the willingness
to kind of actually follow thedata.
It's never perfect.
(21:21):
Humans have biases and thisisn't easy and it takes really
smart people to determine whensomething is correlated versus
causal, et cetera, et cetera.
You can go on and on.
But if there's a commitmentculturally from the company to
kind of lean into, we want tomove towards automation,
optimization, use of data tomake decisions, you will start
to see a shift.
That's kind of number one.
(21:42):
And then I think you have thesekind of tactical follow-ups,
which are interesting, aroundinvesting in storage of data and
capturing of data, working withdata forward vendors, you know,
that's kind of how we fit in.
But all of that kind of anchorsback to this first piece that
you've got to start with acommitment.
You have to believe this is theway to move the company forward
.
Ted Stank (22:02):
Correct me if I'm wrong, but my experience is that
once you start wading in andgetting kind of up to your waist
, your only limit is your team'sinnovativeness and willingness
to try certain things.
You know the whole cliche offail fast, yeah, but you know
just to say, hey, can we usethis for this and to speed this
(22:22):
up and to make us more accurate.
Is that kind of a right way tolook at it?
Jason Murray (22:27):
Absolutely.
It starts building on itself,right.
You capture one piece of dataand that leads to another.
That leads to how you're makingdecisions here and this is
going to help with thesedecisions here, and it builds on
itself, right, and that'sthat's kind of why I talk about
culture and and I think that'sthe only way to do it you can
definitely overcook it on theplanning side.
People just tend to overthink.
It's not like technologists areimmune to this.
(22:50):
I think two years ago, whenchat GPT blew up, there was 500
predictions of where it wasgoing to go, and most of those
haven't actually happened.
It's gone all sorts of otherdirections, right?
I think it's very hard topredict.
You have to kind of be activeand trying to play with it, but
you wait in, even if the thingyou're doing isn't working.
Specifically, you're nowlearning how to operate in this
(23:12):
environment.
Which is going to capture thenext thing that comes online, et
cetera.
Tom Goldsby (23:16):
Hey Jason, I'm going to be serving on a panel
at the ProMat in Chicago inabout 10 days, looking forward
to that, alongside John Paxson,the CEO of MHI, and they gave me
an advanced copy of the annualreport that MHI generates each
year and they said Gen AI isgreat.
We need to figure out how toleverage it.
But the next wave is agentic AI.
(23:39):
Can you help set the table alittle bit, maybe prepare me for
this panel so I can soundintelligent for that session?
Jason Murray (23:44):
coming up, I think , of agentic AI.
It's a bit of a fuzzydefinition, so I'll just give
you how I kind of think about it.
If you were to replace some jobin your company with an AI bot
or an AI agent.
That's how I kind of thinkabout agentic, right, it's a
little disingenuous in the sensethat maybe an example would be
a customer service rep who canrespond to things.
(24:05):
Another example would besomeone who's kind of constantly
going out and testing thesystem autonomously.
Another one would be someonewho's kind of constantly going
out and testing the systemautonomously.
Another one would be likeextracting data from.
You know, I'm going to go tothese bunch of carriers'
websites and try to extract datathat might be useful in
decisioning.
Or imagine like a person usedto have to do that manually and
this thing is just going andkind of summarizing and pulling
(24:26):
it together and bringing it intothe fray, right.
Those are examples of agenticAI.
I think the concept ofautonomous demons or things
running in software that wentout and did stuff that's
obviously been around.
I think what makes it unique iskind of the notion that they're
able to kind of self-adjust alittle bit more than they have
in the past.
That's what makes itinteresting If I'm getting
customer service requests fromsomeone, I'm actually going to
(24:48):
adjust, you know, try to processthrough and make kind of
adjustments to how I'mresponding in a more autonomous
way than like we have been ableto in the past.
And that's a function of justthe power of these models that
are coming online and theability to kind of reason
through some of that stuff.
Ted Stank (25:04):
Hey, Tom, just to be clear, you're going to be on a
panel with John Paxton, CEO ofMHI, right?
Not John Paxson, formerteammate of Michael Jordan on
the Bulls, who used to swishthose threes all the time.
Tom Goldsby (25:18):
That's helpful.
You know, I think just gettingmy pronunciation of agentic AI
is going to be a starting pointfor me, but that's a great
explanation.
And do you see thosecapabilities making their way
into Shippium and what you haveto offer?
Jason Murray (25:31):
You know, I would say this new wave of AI is
largely about automation.
There's so much pressure aroundreplacing jobs and there's so
much energy around that right,and a lot of it's able to
automate these basic tasks thatpeople had to do before.
If I go back to my originalcomment on just the whole thing
is we're just trying to pull inas much data as possible.
The challenge you have insupply chain currently is, you
(25:54):
know, there's like several tensof thousands of carriers in the
United States alone and somemassive number past that and on
the world right, and so they alldon't have EDI or API access to
them.
The thought of can I go out andextract data from these and
pull that into this decisioningprocess, that's been super
interesting.
You know.
Obviously this was maybe moreon the straddling the line
(26:16):
between what you'd callgenerative and agentic, but the
coding stuff is prettyincredible, right.
I mean that's been both fromthe standpoint of our developer
productivity, how much we'reable to get done with a given
developer.
On top of that, the thinkingabout part of the challenge you
have is just integrating.
When I started Chipium, Ithought we build a product and
then people use it right.
Well, if you're really in thespace.
(26:38):
A huge part of the problem isintegrating with people, and so
we've had a lot of luck withbeing able to simplify that
process, and at least the AI isable to help do this initial
mapping, which speeds up thisprocess.
So I think the more this stuffcomes, you're just going to see
more and more use cases as itcontinues to develop and kind of
(26:58):
build out.
You know we have a very kind ofopen approach to it, like I
will try anything.
Sometimes it doesn't work, andthen you just perhaps wait a
little while until you hearsomething else.
But we try to be very open toeverything, and I think there's
a lot of use cases are appearingas we keep digging into this.
Thomas Deakins (27:14):
Good deal.
Let me ask you one morequestion here real quick.
Look into your crystal ball,jason, and I want you to tell me
, like where is the Shippiumroadmap, based on what you're
hearing from the industry andtalking to your customers and I
know you're heavily involved onthe prospect side and sales
cycles too.
But what's the next evolutionof?
Jason Murray (27:33):
Shippium, we've done a really good job on kind
of what I would call theshipping optimization space.
It centers around understandingthat I think we're planning on
pushing in both directions right.
We just kind of launched ourinitial billing management
product for 3PLs, where nowwe're getting into kind of more
some of the invoicing side andhow they mark stuff up, and this
(27:53):
is actually another example ofa great use of Gentic AI, in the
sense that we can potentiallyhave something that is looking
for anomalies and speed up thatinvoicing process in a much more
automated way than it's done inthe past.
You go the other direction, Ithink.
You kind of head more towardsthe customer experience side,
and this goes back to like wherewe started.
(28:13):
The idea is is that we continueto push up into delivery
promise and driving growththrough the platform, as opposed
to being purely about costoptimization.
And then I think the third thingI kind of mentioned, we've
gotten into simulation and havethese large systems that can
(28:41):
just use the same executionengine to process virtual
transactions to come up with,potentially.
What does this lead to from aplanning perspective?
And so we see ourselves aspushing on all fronts.
In a lot of cases it will meanpartnering with people and
perhaps like, if they're good atsome aspect of it, we're able
to kind of take maybe thetransportation piece and push
that into their world also.
But all systems go in terms of.
There just seems to be a lot ofrunway in all directions.
Tom Goldsby (29:03):
That's awesome.
You know, and Ted, this had tocross your mind as Jason was
describing the value propositionof applying AI to complex
transportation problems.
Remember that paper we wroteway back in the year 2000, 25
years ago, yeah, and it was justa concept of all the cool
things we were hoping to do.
Is the future now?
(29:24):
Are we living in that time?
Ted Stank (29:26):
You know, I think like everything, tom, it's
evolving.
Ten years from now there'llprobably be other technology
that can come to play, but youknow, the concepts of supply
chain management haven't changedtremendously.
You know, in a couple ofmillennium probably, but the
technology we apply to make itmore efficient, more effective,
certainly has and will continueto.
So I think one of the reallycool things that I've seen in my
(29:52):
career is the number of peoplewith Jason's kind of background,
this solid computer engineeringbackground, coming into the
supply chain world.
Like I'll have students come upto me and say I'm a supply
chain major but I'm thinkingabout doing business analytics
as a minor.
Do you think that's a goodthing?
I'm like, dude, you want to geta good six figure opening
salary coming out of theundergrad program and, yeah, do
that.
That's a really good thing.
(30:14):
Not to pick on my marketingcolleagues, but so many of our
students for a long time weresupply chain majors, marketing
minors, and I always tell themwell, you have a good career if
you want to go into sales with a3PL, but if you really want to
roll up your sleeves and getinvolved with where the world is
going, it's really with thedata side, the other thing I'd
like to highlight.
Jason, you mentioned culture.
(30:34):
So you're not just pureengineering anymore.
You've learned right.
I have my roots in engineering,my roots in operations, and you
know, I used to just thinkculture was this bogus stuff
that psychologists talked about.
But man, culture is everything100%.
Jason Murray (30:49):
As Drucker said, culture eats strategy.
Right, a lot, you know, andwe're a long ways away from
people not being involved.
You need the motivation and thepush and all the other stuff,
and that's a critical part of anorganization is anything.
You can have all the technologyyou want, but if it's not being
used or embraced in the rightway, it won't go anywhere.
You see that over and over.
Ted Stank (31:10):
That's why it's all about leadership, right.
Jason Murray (31:12):
Yeah, exactly.
Tom Goldsby (31:14):
Well, I think we got to wrap up about now.
But hey, Jason, thank you somuch.
We Well, I think we got to wrapup about now, but hey, Jason,
thank you so much.
We've been.
You know this has been a datethat's been circled on our
calendar for some time andthings are changing so rapidly.
So thank you very much forsorting things out for us and
giving a sense for where we are,the potential of AI and also
where it's heading.
And hey, Thomas, thank you verymuch for joining us.
As well as a co-host, you wereinstrumental in bringing Jason
(31:38):
into our fold.
All as a co-host, you wereinstrumental in bringing Jason
into our fold, so we reallyappreciate you for that and
again, elevating our level ofplay, just as our team around
GSCI does each and every day.
And hey, just to give a littlebit of a teaser of what's in
store, the next supply chainforum is coming up in early
April and, Ted, I think we'vegot a special guest.
Colin Yankee, the Chief SupplyChain Officer of Tractor Supply,
(32:00):
is going to be our specialguest on the next podcast.
Right?
Ted Stank (32:03):
Absolutely.
We'll be doing our next podcastlive from the forum on April
9th, I think.
Is that right, Thomas?
I think so.
April 9th with Colin Yankeefrom Tractor Supply.
Tom Goldsby (32:13):
Another great partner of ours, Jason.
Ted Stank (32:14):
I echo my thanks to you and to Thomas.
Great to have you guys with us.
Wish we were all togetherbecause I would go out and buy a
round for us all.
Jason Murray (32:23):
Oh it sounds good.
Ted Stank (32:24):
Maybe you can do it and put in a request on Venmo
for me to pay you for it.
Tom Goldsby (32:29):
Yes.
No pappy though no pappy Topshelf, jason, you earned it, man
, all right everybody.
Ted Stank (32:35):
Thanks, as always.
If anybody has any questions,suggestions, you can reach us at
gsciutkedu Until next time bewell.
Intro & Outro (32:44):
Thanks for tuning in to Tennessee on Supply Chain
Management.
If you enjoyed the episode,subscribe today on your favorite
listening platform to get allof our episodes as soon as they
drop, and don't forget to take amoment to leave us a rating.
Have any questions, thoughts orfeedback?
We'd love to hear from ourlisteners.
Email us at gsci at utkedu.
Join us next time as wecontinue pulling back the
(33:07):
curtain on the world of supplychain, educating and
entertaining you along the way.
Until then, listeners.
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
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!