The $600M Freight Founder: Most Forwarders Won't Survive the AI Shift

What does AI bring to the table?

Yeah, AI is becoming very intelligent.

Who is leading the way in AI adoption in your view?

Companies where the leadership has understood the importance of AI.

Would you rebuild a pure technology company? Would you build a competitor CargoWise 10 years ago?

I'm asking myself that question almost every night where I can't sleep. And then you know I remember a board member, one of our board members said why don't you guys just do Shanghai to Hamburg like as a trade lane just do Shanghai to Hamburg

you think it's that far away

question is what is faster when you work on a freight forwarder today and I see that for all of our customers around the world many say I have real problems finding great operators. Welcome to the supply chain technology podcast. I'm very happy to have today Erik Muttersbach.

My German accent is

it's perfect.

Is kind of uh kind of okay. Uh Eric, welcome. Great to see you. How are you doing?

Thank you. Doing well. First time in Miami. And um I realize I have to uh beef up my Spanish.

Yes. Yes. Miami you need to have some Spanish definitely for taxes and for and for multiple purpose. Yeah. I actually often time with my look people speak to me in Spanish first in most of the instances actually. So, so um Eric, it's great to have you. Uh AI is changing a lot of things in supply chain. We'll talk about this. We'll talk about you. Um Eric, serial entrepreneur and uh building now Zauber, CEO and founder of Zauber. So, uh Eric, maybe let's start with uh let's jump into it. Let's start with u tell us about you maybe first and uh and Zauber.

Yeah, thank you. Um thank you for having me. No. Um, look, well, my background originally is on the computer science side. I always loved building stuff. Um, and um, ended up u in computer science somehow by accident. Um and like I got drawn into being one of the first hires, founding engineer, founding CTO later, um of a company called Daily Deal, um which is a which was a Groupon on copycat, I think it's fair to say, in Germany in 2009. Um and uh

I was uh sort of still studying, but I worked at that company and built up the tech and hired the first team together with Fabian and Ferry, the founders. And um, of course it was absolutely naive but they ended up being able to sell the company to Google I think two years into the business and it was just a fantastic ride um you know it was so interesting um to see what we would be able to do with a small team and it was for me really the entry into the startup world and created a lot of excitement for entrepreneurship and that that's what I ended up doing more in life afterwards. So um I subsequently um founded a business which was um uh I think called cloud numbers. So today you would say uh I think um maybe uh Vercel for Jupyter notebooks but in 2011 and 12 was called Cloud Numbers. Um which worked out okay. I think um you know we we made a few mistakes but we had a had a really nice product. Um um and we I moved on to build a bootstrapped app business. Um and then in 2015 ultimately started a company called Forto which is um a tech um freight forwarding company. Um I had not been in logistics before but my father actually was a sailor on Hapag-Lloyd ships. Um Oh wow.

Interesting. his father. So my grandfather um is been an engineer on the GDR shipyard in East German shipyard building one of the first um container ships. I mean um I think multi-purpose ships which could also carry containers for the East German fleet. Um and somehow I love sailing, I love economics and I found this industry always interesting. And so after doing B2C and some um some more like tech product like developer focused products um I um looking into different industries I felt drawn to logistics. I felt drawn to supply chain. Um I just thought it was I was just naturally curious about the industry. And obviously if you look at logistics, you know, um it's this huge trillion dollar industry that moves the world quite literally. And in 2015, it looked to me like um it moves the world and software also has moved the world, but somehow software hadn't moved logistics a lot or at all almost um especially not what I was looking into at that time, which was freight forwarding. Um and so I felt hey um we should that there should be like a huge opportunity for logist for technology and logistics and like better service higher productivity. I couldn't verbalize it at that time like that, but better revenue management. You know, today we'd say as as freight forwarders or as logistics companies, be on top of every invoice, be on top of every de merge attention case on every change of the shipment. And um so we said with our naive sort of um 25 year old selves, my co-work um

you were 25 at the time when you studied Forto.

Oh yeah. Um and we said, "Hey, come on. It can't be that hard to build a freight for water. We will just build a freight forwarding company." Um and we went and bought these books that you do when that you that you buy first year when you do an apprenticeship in Germany about what's the bill of lading. Uh one of the questions on the one of the first pages I remember was um when you cross when you go with a truck from Hamburg to Munich which highway joints do you cross? So, so all these things did not at all prepare us for what was going to come. But so we started a freight forwarding company and we said we will just start the thread forwarding company and we will um build all the tech ourselves and our our own TMS our own customer front end with full self-service capabilities and so on our own rate management and um we think that we can provide great customer service because you can self-s serve yourself we'll do great track and trace um and we'll significantly automate the business. So we started in 2015 for the company is still around. Um I ran all the technology stuff there as a CPO CTO. Um and um I did it for almost 10 years. Moved on a few years back and handed over all my uh operational responsibilities. But um when I when I had moved on, we were close to a thousand people, 20 offices in Asia and Europe, mostly European controlled seafr air freight. Um and it was really a fantastic time. I mean so many great learnings. Um um and uh I mean in many ways um I'm very grateful for what we were I think able to do but one of the learnings also was you know looking back at the initial hypothesis of we can with software significantly automate the business is software just doesn't cut it for logistics at least for freight forwarding I would say software doesn't cut it

um from the outside

why why is that let's get exactly into that

from the outside logistics looks like this. Um I'm going to quote Pavle from 0.9 as one of our investors. Um not not Pavl directly but 0.9 as a fund and he said um logistics looks like this beautiful trillion dollar industry that you could absolutely easily streamline with software and now maybe AI but if you look into it you have so many details so many exceptions nothing really works based on a blueprint and I think that's really what we discovered. So I think two core problems that we that we found out. The first thing is um at the time where you only had software only quote unquote um so 2015 those those years obviously the pattern was let's try to um structure all the data in some relational form. Um that means that we need to make customers stop sending emails for a booking but make them go to our platform and place a booking there so that we have the data in a structured format and then we can move on with sort of software style automation. if then else and code the workflows that you have to take after booking maybe get back if data is missing make the carrier booking etc. Um the reality is I think you cannot get stakeholders generally not only customers but also haulage companies etc. You cannot move people stakeholders away from email. You cannot get them to stop asking for um an Excel sheet on Friday with deliveries that come next week into their warehouse. um you we never found a way for the bigger accounts, you know, we ultimately moved also to bigger accounts. Of course, we you could we could never find a way to encode all the details of a customer SOP or a work instruction like a specific way to handle a specific client. For example, what what rules do you find there? You handle a bigger automotive client and then you have um an example from this week. um you have requirements in there that for every um document you send us, you must reference the lane ID somewhere in the document and then before the shipments arrive um for all the long beachbound shipments, you must send us seven days before and inbound to um inbound us at this company.com and if there is a two days delay in the transit time, then you must alert us with three backup options, etc. And um the reality is that um these details we could never find a way to put them into a a sort of um a structured data format. You know all this this long tale of individual customer requirements and that was just really really hard. So the complexity of unstructured data and being unable to put them into that into a structured format that was I think a big topic and the other one was that even in the scenarios where we had or in the parts of the value chain where we were able to structure the data in a table or let's say put it into our own TMS well um we we realized that software isn't good enough to do the automation. um you know um uh why because a delivery of a container around Hamburg is already quite different than Berlin my in my example and Berlin is just um in US terms 120 miles away you know but it's it's really different and you just um you start the whole process quite differently you maybe work even with different companies etc um and so there's just um so much nuance decision-making that we just found that you can't encode that in software flows um and I think the proof is um when When I speak to customers today, all of us, we tried RPA.

Yeah.

Um and I think it works for your top 10 clients or your top 20 or so depending how big you are, top 50 where let's say the transport order always looks exactly the same. But um it doesn't really scale. And I think you know if I if you look at the um at the public reports of the of the companies that are public and um I don't think I think we see that the productivity in terms of file per person the last 10 years they haven't moved that much. Yeah.

Um so I think it's a general theme in the industry.

Yeah. So then software software has not so um let me go back um a little bit. You said you said something I would love to jump before we go into ZA and the realization of like software is not enough. Right. I think this a great point and I also have a a view on this. Um you talked about it cannot be that hard to build a freight forward one right it is so I come from the operator world I come from more the traditional world right at DHL um

I had always the impression that a lot of tech founders

did underestimate the complexity of building a business in this space

100%

what what was your learning there and so maybe for also for future founders or people that have you know I mean you build a thousand people company, right?

Um, by the way, for the audience that doesn't know Forto, but I don't know if it's fair for me to say that, but I think it's a pretty good analogy, but you know, it's a a business model similar to flex sport or now port, which is a digital freight forwarder as some investors like to call it um through stack technology, but providing technology and the service itself. Is that is that fair?

Yeah. Yeah.

Great. So, so just so for people, so tell us about that. It cannot be that hard to build a freight forwarder.

Yeah. Well, um,

these people are stupid. They're old school.

I have a much more nuanced perspective on building tech enabled service businesses than probably the VC community on average. Okay.

I think is the biggest hype in VC at the moment to say or like from investors or to say, "Yeah, let's let's sort of not just build AI products, but let's build the whole business because then we can capture sort of the full revenue." I think what ends up happening is you lose yourself in the operational details um of a um of a service business, you know. Um, why was it hard to build that in logistics? First of all, I think you know it is it was really hard to find people

who were really experienced logistics sort of freight forwarding professionals

um that the same time would be willing to engage in an early stage startup. You know, we had people that literally left two times after 3 days because they arrived at the company and they came from a big carrier or very big freight forwarder from Hamburg and um they saw the sophistication of how we worked and they got so scared they moved out um after 3 days and went back to their old jobs.

They were looking for the fax machine

maybe. Um and I think that's okay, you know, it's okay. It's just especially 2015 startups especially maybe in Europe was much less of a sort of a well-known career path maybe especially if you come from logistics. So it was just really hard to find people um who we could bring on board. Um partly because we were thinking sort of very tech company style and the people were not at all used to tech company style thinking. So that means you know really focused lots of I don't know Excel sheets lots of slideshows whereas maybe in logistics lots of live knowledge lot sometimes a bit more gut feel obviously also there's people that do very analytical thinking but you know um I think you know we were just I don't know I don't know I just perceived the day-to-day way of that we work was very different and that often clashes between cultures of those different groups and so we had a really hard time bringing people in that understand the industry well which was a big problem for us primarily and I think we should have put sort of with regards a recommendation to future founders. We should have put a much bigger effort into trying to bridge that gap. I think today the one thing that I would have done much much earlier is to focus much harder on finding the right people from the industry um who are senior enough to structurally understand the industry um and um would help us would have helped us to um sort of um I think make the business successful and I think you know you need to that is that means sort of I think a lot of hard work and a lot of triangulation who are the right people but also I think building a culture where those people can can be successful. Um the the second big topic is I think that um uh I think we should we probably also aimed at one of our mistakes I think was aimed at rebuilding the entire technology for a freight forwarder a toz from the get-go and I think we should have picked the right battles. Um I think you know in the beginning

did you like did you build CRM as well and so

no CRM not but we built a TMS which is a feat in itself right I mean I think we once reverse engineered I would think I would think CargoWise probably

has spent 800 million or so US dollars or 600 or 400 with that order of magnitude to build a TMS that's really obviously fairly capable but

yeah probably even more right because if you look at R&Ds of a 20 years timeline right

100% I mean if you were like smart and you could rebuild it today probably you you can even do it in 50 to 100. But um I think um it just takes a huge amount of money to build a very capable TMS and you need a fairly from the beginning a capable TMS because you're immediately launching in multiple countries, multiple um sort of regulatory regimes, etc.

But this an interesting one, right? Because like let me let me just like jump that one because so um Augur

Yeah.

raised 100 million. Yes.

Three years ago. Yes. Right. With I think a promise of we're going to rebuild that.

So they have 100 million in the bank. They have extremely smart people, extremely capable people.

We are three years in. We haven't seen yet the outcome, right? So, so can that number be challenged like the 50 200 million you can rebuild that?

I think Augur is also BCO and shipper focused, right? Yeah. Look, I I think um

because I mean the reason why I want to because I think I see so many people like simplifying this like yeah you can rebuild this CargoWise is [ __ ] you know all this like

yeah% maybe you can criticize maybe 20 years software and so on but

to me it's like I look at people like Dave Clark I'm like if they have not been able yet to pull out this thing knowing how how how you know how solid is that guy I mean to me he's one of the best guy in our industry for Yeah,

I I'm like, well, they probably going to pull it off at some point, but it's still, it's it's already three years. They was announced the 100 million.

Lots of respect for uh sort of I think putting such a team together and raising so much money and pursuing such a big dream. I've saw I saw their booth at um Manifest and I think they have very good um let's say opinionated um communication. Uh it is a little bit maybe that's my European me, but I think it's a bit fishy that after three years you don't see something concrete. Um there's not your German pragmatic approach.

100%. Maybe that's as that is that. No, but I do think you know, look, let me put it like this. For me, it's not the right approach anymore. Um I really um I think you know I'm I really focus on creating value along the way because I also want to make sure that the journey I'm on is the right one. Um and I think you know in the industries that we are in um it should be possible um to aim for that. I think you know if you're building rockets and fly to Mars I think it's okay to not show something for 10 years. I think in the age of AI with AI coding, if you're building software for logistics, um there should be something along the way. Maybe they're just not talking about it, but in the way that I would build a company is, you know, create value quickly and like um also learn from customers and be really really customer-centric. And I think that mandates you to have something out early even if it's not fully

Yeah. Yeah. No, absolutely. And so you um you mentioned software was not enough.

Yeah.

Right. And so what does AI bring to the table?

Yeah. that software could not do before and that becomes that game-changing.

Yeah.

Into the industry.

In my mental model, the two key changes here are or the two key unlocks that AI brings are exactly connected to the two challenges I we were facing. Challenge number one was all the unstructured data and it didn't make sense to upgrade people from email to a portal. Um it doesn't make sense for them. Also, the portals were never good enough. So, it just didn't work. Also on the other side, we said decision-making um and the ability to encode the decision-making that great operations clerks have in their head today in software just wasn't possible. It wasn't technically not feasible. Um I think both of these things are now addressed by AI. Why? Obviously, I think we all understand that now is AI is fantastic at processing unstructured data. Um it can look into emails, Excel sheets, word files. It can now look at websites and um it can also communicate with customers um shippers whatever via email and um perfectly sort of I think navigate this unstructured world. Um that's a huge unlock in an industry where I also think email will prevail. Um then um on the other side AI if you look at the benchmarks I think is super impressive how much it has moved the last three years. AI is um becoming very very intelligent. um and is able to mimic a lot of human decision-making it's able to work off natural language written um process instructions and so in a much much better way and by now very reliable way you can encode um processes um that humans read through before and were able to sort of perfectly get and execute now an AI will do the same I think it was um was not possible with software before and I think that's the huge unlock and why I think

I think we're now at a zero to one moment for um how logistics will operate like many other industries.

Yeah. So I think my my view was was also on I think you um you touched on it a bit earlier but my view is you have even if you have an organization like let's say Forto that is highly technologically advanced.

Yeah.

You still have to interact and you still have to exchange with a lot of stakeholders.

Yeah. And to me that was always the challenge around like how much a software by itself can increase the productivity because

if I'm dealing with a shipper that still works on email and on phone calls well

my productivity still gets tied to that process. Um what I do see with AI is that yes now you can have even if you have a software a traditional software yeah as a core of your operating system and now all this interaction with the outside world

shipping line airline shippers final clients

you know whatever infrastructure partners you have customs and so on you can now build agents that actually automate a lot of this is that is that the way you see it as well

100% 100% see it like that I personally think that email is the fabric of the future because um if we had to agree on a different format or protocol, I think you know that's already by definition than a no-go. We've tried in the how many standardization initiatives have there been in logistics that always have failed, right? And in the end, everyone still sends emails or like somewhat oddly formatted Excel sheets around and I think, you know, now AI can just do it. So you can easily plug in AI into the way that we work today. And so I definitely think for the next 5 to 10 years, you'll see AI communicate to AI or to a human over email. Um we see that today actually that um in the very sort of infancy stages one in I don't know 100 a thousand emails where we mostly work for freight forers and with Zauber now and we work for freight forwarders and one of the use cases we do is spot quote automation and um shippers um or like let's say controlling customers for that sense um they also start obviously to do deploy AI and especially for um spot uh quote requests uh it seems that some of these tendering platforms or um quotation management platforms that the shipper side uses, they can trigger emails with AI to freight forwarders and once in a while we see that u our agents respond to these because there is like a footer sent by AI or something like that.

Um and so um today often there's a human in the loop before our our agents are able and able to send out something but in the future you can very much imagine that AI talks to AI um via email. I think I don't want to

How far is that or is it already happening?

It's already happening. Yeah, that's what I mean. It's already happening. Um but it's more like I think one in a 100 or

Yeah,

I don't even have the number of matter. I justve I've seen a few examples like that in our in our data. But um just

Yeah, I mean just just scratching the surface. But is it fair to say that in general anyway AI in supply chain is just scratching the surface?

Oh, you scratching the surface?

Yeah, we are just at the beginning right of the journey.

Yeah, we're definitely at the beginning 100%. Yeah, we're definitely at the beginning despite um bold claims from marketing teams and so on. Yeah.

Yeah. I mean um I mean in the enterprise world

you probably have less than 1% of adoption even in the largest companies around like potential workflows automated. Yeah.

I mean this 2026 you know

there's a few leaders but I that maybe are even already beyond that. Um

but as an enterprise or as a one workflow, right? Because you do have you hear like the likes of CDH Robinson and companies like that talk talk about like wider adoption

uh but their business model you know includes a lot of type of service so maybe they have a high level of automation maybe in trucking

domestically in the US but then when you look into the freight forwarding aspect maybe it's a bit less

yeah um I think definitely trucking is much more advanced freight forwarding is less I think the processes are a bit more complex um and there's also less companies I think working in that field at the moment from a supplier side. I would say that if you have a company that's embraced what the technology can already do today, they've probably already seen a 7 and a half% and I'm talking about sea and air freight 7 and a half% productivity increase on a file per person basis. So I think that's already quite a meaningful step um compared to everything else we had before. I don't think you know it's by far not the end and uh everyone will will try to drive much more and I think much more is possible but I think if you fully embraced it let's say on in a few countries I think you have probably been able to see that

okay okay and um who do you think is leading the way

who

who is leading the way in AI adoption in your view

who's leading the way um

if you don't want to mention names maybe um like trucking maybe trucking is maybe a bit more

just think about this so I think definitely trucking um who is leading the way from a customer architect perspective I think is the following companies. Companies where the leadership has understood the importance of AI and um of um just trying it without academic ROI. um obviously you know do your structured vendor selection and so on but without an academic purchasing or procurement project that runs over 12 months to um do something where the outcome might be unclear for the people involved at the moment. So companies who are willing to sort of just try things out um of course with with with a structured process behind it but you know to go a bit faster. Um I think also the companies where I think they are most successful also with deployments the leadership CEO has a strong narrative around how what AI means for them and what how AI will make them stronger and how AI will change their business and they repeat it again and again and again. Um uh I think that's super important because there's obviously the biggest topic is change management and adoption. you know technology is already much much further than I don't know 5% 7% 10% productivity improvement if you if we were to like focus on that KPI um but I think change management is the biggest one so the especially when people have a business like logistics you have to really think about how do I take the people with me um the other components are um you have a very capable um I think um deployments and actual change management implementation process and team um I see that companies have more recently for example, deployed in our industry CargoWise. Um, they have just recently had they just recently had to figure out how do I deploy a new piece of technology into all my branches make people adopt a certain new flow. And I think these companies have a benefit because they have that muscle of deploying technology um around the world. And then last component, the third component is I think a strong opex BPM operational excellence um BPM process team with a lot of clarity on how do we actually want to do it? How do we roughly do it today? Because you know AI is really smart but you need to still customize it to a degree to your workflows. What do you want to achieve with this? And if in the end you let every branch manager decide how they will roll out the AI and how they will do it, I think it's just more messy and it will take much more time to adopt it. So I think these three things are super important. I think conviction and the strong narrative um and a bit of a sandboxing approach from the from the top level management. um I think u a great muscle and capability around change management and deploying new technology and um I think a strong process excellence

in ter yeah yeah so you touch on something that I think is really also interesting because there's a big debate on this so um many people say well the is not there I mean I spoke to just earlier this week with a very large consumer electronic um head of innovation uh company and he was just telling me like even the they're trying to build it internally and he was talking about the cost or the cost is even prohibitive. It makes more sense to do it with people.

I was like I was very surprised by that comment. But the point I wanted to go back is

you mentioned

you might not need to have or you should be able to move even though you do not have a very clear ROI and I agree fully with that statement. I think it is a new space

and trying to adopt this in the same way you've done RFPs

for WMS or for traditional TMS that are systems that you know exactly what are the criterias you know exactly how to deploy you know to compare cost

and and trying to do the exact same process to deploy AI think is a big mistake is that kind of

I agree with that for um last year and I agree with that probably for the next 6 to 12 months but I also think that we also have to hold us accountable to making sure that it makes sense for the business ROI is the mathematical way of saying it, right? But I think ultimately um the thing has to has to graduate from um it's an interesting technology to it really fundamentally impacts my business. Um that is always the case. I think for blockchain and predictive AI and ML and for example freight forwarding that was never really the case. You know, it didn't move the needle and then you should get rid of it. But I mean I mean I don't know for you but there's no doubt right that I mean at least at least and I I have I have strong conviction that you probably don't have I mean you're building in this space but there's no doubt

that this going to work out right the question the question is how quick

can you get there I mean the that the ROI is going to be there that I mean to me it's more around like finding the right use case finding the right place having the right change management it's more around where do you do it then if you should do it is is that

I do think that we live at my I'm citing my wife here who's um also in that space and um

she works in the AI too

she's until yesterday been a GP at a pre-seed fund in Europe at visionaries and now

did she invest in the company

um no no we don't do we don't do that in and for she worked on a fund that was invested in my

okay

in but um and now since since yesterday she's um CEO of Langdock um but um

coy But uh I'll borrow from her um a quote. We are living currently in the holy [ __ ] And the holy [ __ ] is that there is some people who have one of our customers one of the board members has tried Claude and is a heavy Claude Code user I would say by now and um I think that created a conviction next to his him generally being super tech heavy and very innovative and I think very visionary uh and deeply rooted in the business as well. I think that created this deep conviction uh by you using Claude Code. I think um you know he basically took some data from his data warehouse and built like a huge simulation for how some parts of his company works and uh he showed me and it's super impressive. um you know it's like I don't know in the past engines would have worked months for that and now he can go to some of his branch managers and um say hey look I looked at your data and that's like is a like a it's a trucking part of the business and you know um that's how uh that's how much loads are coming in that's how many loads go out that's how sort of some of the cross-docking operations work can we optimize this or that I think you know that's that's how he told me he's using it and anyway I think this the best case because if you are if you have tried Claude Code and you've built something with This as an example it could be lovable or replet you are mind blown and there is almost no question anymore that this will change the world and you live with a constant fear of how's the future going to look like damn I need to transform my business and also man we need to roll this out really really quickly because it's going to be so transformative on the other hand if you have not tried it um then you then you think oh maybe there's another next blockchain next um I don't know next ML everyone talks about it and it doesn't make sense and you by far not as convinced but it's incredibly hard to bridge that holy [ __ ] gap between the people who say holy [ __ ] this going to change the world and who think holy [ __ ] what are these guys talking actually about I mean this a massive hype I mean the bloombox said the biggest sort of scam in America currently is um chip companies and AI companies shoving money between each other um which doesn't exist yet right and driving up the valuations but um but I think uh you need to be true at the same time

can be true at the same time um but I think you know in order I think everyone should who has sort of a leadership position at the moment needs to bridge that holy [ __ ] gap for them. Yeah. And get from holy [ __ ] this a hype to holy [ __ ] I'm I'm convinced or maybe also not but I just have not seen any single person who doesn't say this going to be fundamentally a fundamental gamecher for our industry in logistics for example. Yeah, I love it. I love the holy [ __ ] moment because I mean many people talk about the chat GPT moment, right? And I think there's a lot of questions as to the ChatGPT moment was probably the easiest way for a wide number of people to test

what's what's possible and what's the potential. But I think you're spot on and when you say

yeah, what about the second ChatGPT moment which is now the holy-fuck moment. We we coined that uh with your with your wife. She coined she coined it. Uh but the um the how do you do that right like so for executives in companies because I I'm fully with you I think there's a lot of executives and I was in a lot of these boardrooms that

they understand but they don't experiment

and there's a huge difference between both right a

Uber is fantastic

you know you have access to cars

but when a car comes in front of the house pick you up, you know that you schedule it, you don't have to pay, you don't have to put a card, you don't have to

it's this an old school example, right? But this really what happen when you actually experiment versus you understand what what is possible.

So how do you do that? How do you get this exact do you put them into a room and you tell them you know what you John or Jack or Pedro or Juan you're gonna today you know you're 60 years old but we're going to help you you're going to build an app in 30 minutes. Is that the way to do?

Um, it's some it's in going into that direction. So, I'm not I never really like that anyway. Maybe because of computer science background, but when I talk to a customer first time, for example, to a board member or sea level or senior leader, I never show slides. Um, I and I jump right away into a demo. First of all, because everyone's pitching AI and, you know, I think, you know, it's important to make it haptic and concrete and I always jump right into a demo. And um in addition I let people sort of take the role of the counterparty. So for example one of our demos um is um an import coordination uh process where the shipment arrives and then you have to talk to the the receiver the consignee um about when do you want to have the shipment and my the AI proposes the communication we send to the to the to the consignee. And then I let my counterpart like let's say the customer here I say hey you're now the you're now in the warehouse inbound team. what do you want to say? And my AI proposes normally to say delivery between Wednesday and Friday, 7 a.m. to 3 p.m. because outside it's like not demurrage-free anymore. Um, and then typically the people respond and say, "Yeah, I don't know. Then let's do Thursday, you know, in the middle. Let's do the same thing." I say, "No, no, no. Make it harder. Make it harder." You know, and the he says, "No, normally we end up with something like um yeah, you know, I need to have the delivery." says then sort of the person who models the inbound warehouse team on Tuesday uh at 9:00 a.m. And then the AI comes back and in a smart way react and says, "Hey, okay, then we might have demurrage fees of that size. Can you confirm with that?" And so you show a little bit some of these exceptions that are now able to automate. And you show how well it's integrated into emails and you let people participate a little bit um in it. And that sort of maybe as a GTM recipe almost um that works for me well. It maybe is it's like 1% of what I mean with seeing it for yourself. But then um when I give a talk, my my last there I have a slide. My last thought as always, if you take one thing away from today about AI in general or so is by tonight you have have to have tried either Claude if you feel that you're technical and you want to swim up a terminal or maybe lovable on the other end of the spectrum or maybe Replet if you're in the middle. Um but you have to try this you know if you're a responsible leader you have to try it.

Yeah. Yeah. I think I think you you are really um you're really spot on on this. You have to try it. You have to try it. And uh you know I mean I think even the simpler example is

if you have any type of like Excel document that you review

just just if there's no confidential information or just remove the confidential information and just just put it there like a database or whatever and say analyze this thing and you you'll be blown away right I mean you'll be blown away just by a simple a simple analysis of data or dashboarding or you could literally you know you can really do like simple things

that might create really a bit that holy-fuck moment. Yeah, definitely

for these executives and these teams. Um

tell us um a bit more about the company today. So there's obviously a lot of companies there's a lot of a hype around AI in supply chain.

U you are tackling I think you talked about the freight forwarders uh as one of the core area um of the business.

What do you do and how do you do it? Right? You talk about sales quotation type of automation. What does Zauber do really in in the more detailed approach?

Yes. Um, so my my conviction is that 80% of what we do today manually will be aided by or run by AI in the future in freight forwarding.

80%

80% from spot quotation to invoice reconciliation and accounting.

Mhm. Um and um the residual 20% I think is um really longtail crazy exceptions where humans and I think we already see that today as a pattern are still sometimes a little bit more flexible um is a little bit more versatile in this multimodal world um and um I think the 20% um that humans still do are either like crazy difficult longtail high stakes exception handling and I think relationship um uh with customers um as long as there is a customer a human customer on the other side. I think you have an AI customer on the other side which we will I'm I'm convinced by the way that we will have in a year from now or maybe already earlier completely AI run um e-commerce jobs and then in that case uh in that scenario I think you know um there's no nothing that speaks against like end to end AI run um logistics op right anyway let's go back to what we do today so I think 80% of the stuff will be automated and um we today u basically provide

email me SMS voice like across

yeah also like the workflow I would say multi-channel all the channels that you have um data entry decision making um proactive checks all these things um so the end to- end workflow um and I think that um and our goal basically is as to um help companies with um sort of having an AI workforce that um executes this um we are very focused today on um sea and air freight um automation um with AI um uh this my background. This also a little bit my chip on the shoulder. I really want to solve that. You know, I've tried 10 years and I think now we have the technology to do that. Um and so very specifically today sort of the bread and butter use cases are spot quote automation. Um that is obviously also a productivity topic. Um u but also very often you know it's about turnaround time to customers especially in air freight. Um when it's expedited freight um you just very often just need to get a quote around the corner very quickly um for the customer to accept it and it matters that you that you're fast. Um, also very often, you know, in the rates world, you don't have all the rates at hand. So, you need to reach out to your agents overseas and when you wake up, let's say you work in Miami. Um, you wake up, you probably have, you're in the pricing inside pricing team, you have 30, 40 requests already from your branches or customers in Europe. Um and um now uh you know bridging those time zone gaps I think is perfectly great opportunity for AI because AI can already go while you're sleeping and acquire some truckers or maybe some other overseas offices that you need for some local charges from Vietnam or the West Coast or something like that. Um so the other use cases are a little bit more on the operational side. Um file creation, file updates, um coordination with customers, all kinds of data entry jobs. So more the simple stuff that I think you know gets you from zero to um 0.2 and starts the AI journey a little bit. Um that is sort of what we do today. Um growing into sort of as models get better as customers also I think learn to understand how to use the technology obviously as our product also matures uh into more complicated use cases now like um allocation management uh like import um haulage and delivery coordination um and so on. Why why focusing on the freight forwarding industry?

Um I think there is um an emotional topic connected to it which is which is I I did it 10 years and I really wanted to nail this and I left for with a lot of gratitude and I think also a little bit of pride of what we were able to achieve despite not everything we wanted to do came true. Um but also with a as I said a chip on the shoulder I that I pains me a little bit that particularly for me who's been on the product and technology side of the company um I think we weren't able to create that zero to one moment and that we were hoping for and I think now we can so I it's a bit become maybe mission for life to do that. Do you think anybody has created that zero to one in freight forwarding in our industry?

I think CargoWise did that you know 20 years ago or something like that. I think clearly um they did something right and they provided an important foundational layer for the for the for the industry but I think now is next step and next generation with regards to the next value creation I think will come from EI AI.

Yeah. Do you think flexport moved the needle in the industry? I do think that now ports, flexport and forto as sort of I think the three main digital forwarders um are in a great position um to leverage AI because um one of the biggest bottlenecks that I observe today is not so much anymore the technology but maybe the conviction of the leadership, the ability to change manage the organization um and to have process clarity and I think these organizations um are in a prime position to have all of these things in place. So I think they're in a good position. I do also think, however, that there is still a lot of manual things that distract you. You know, one of the reasons why I said this time with Zauber, I want to be a technology company and not an ops business anymore is because, well, first of all, I have a computer science background, so maybe there's a 5% more excitement than that. But um I remember one of the board meetings where I was about to present um a 45-minute um tech strategy, AI, still heavily ML stuff at that time, ML, AI strategy and so on. and uh our chief of staff, Adrian, at that time came to me and said, "Eric, um you got to shorten it to 10 minutes or 15 minutes or something like that." I said, "Why?" I said, "Yeah, well, one of our most experienced operations managers um he messed up um and we um have had a few hundred containers um left sitting in Hamburg, I think it was, um for I don't know weeks." And so we had a seven figure demurrage bill coming in and now obviously we have to talk about like first of all that happened, how we prevent that in the future, etc., etc. And I think it's a good example for how much running a service business to my earlier point can distract you from being fully focused on driving technology forward. So that was my deliberate decision to say I will be a tech company today. Um but I do think you know there are certain organizations that are in a good position to leverage technology.

Yeah. I mean at least on the change management and the value of technology the leadership understanding technology as a game-changing for sure. For sure it's uh it's it's very relevant. Um, if you had to rebuild Forto because as you talked about like service versus pro technology,

would you rebuild a pure technology company? Would you build a competitor CargoWise 10 years ago?

And if that would have been successful, what would have been the potential there?

I'm asking myself that question almost every night where I can't sleep.

Okay.

Luckily, I can really sleep well these days because we have kids and every single second that are really young. Every single second I can

How old are they? How old? one and three. So, every time um we can sleep and we're not interrupted by I do sleep. Um but

Okay. Yes. Yes.

No, look um we've been asking that ourselves a lot. I do think um it was definitely and for it wasn't unsuccessful, but I think you know it wasn't that zero to one moment that we were hoping for clearly um because the tech wasn't ready. Um and I think it was a something worth having tried. Very worth having tried. I think that that time that the company was built the two paths would have been maybe be more of a freight forwarder and less of a tech company and embrace you know what a tech company is all about. Let's be super focused. let's do one thing and let's do it really really well right and let's put like great technology behind it and um and then you know I remember a board me one of board members said why don't you guys just do Shanghai to Hamburg like as a trade lane just do Shanghai to Hamburg um and that's very much like the tech investor thinking maybe that and I think not no not no bad intent there but that clearly doesn't work as a freight forwarder no one does Shanghai Hamburg right

you're not afraid for if you do one lane

a good freight forwarder will say Shanghai Hamburg ah you also do Timbuktu to um southern Antarctica whatever yeah yeah know we we've got a lot of experience. As the customer sends their quote over, you scramble to get on the phone and find some agents and set things up, you know, um, and if someone has like a coal break shipment from Australia to Hong Kong, sure, you'll do that because, you know, anyway, so

and I can I cannot I cannot I cannot imagine you responding to an RFP with like 200 lanes and you have one lane.

There you go. So I think you know I think there's maybe from today's perspective you should should have maybe made a choice a few years in to say let's be more of a freight forwarder and allow us to not use our own TMS that we've self-built not use um I don't know a platform that can already service this lane etc for our customers but just do the plain freight forwarding and focus on a few very select parts of the value creation of reporter let's say today I'd probably invest much more into things like um allocation management I think it's a huge GP driver that you can more easily roll out um in a in an existing business if you have sort of the clarity and technology um and you didn't need AI for that so much. Yeah.

Um I probably um focused less on trying to build a super holistic TMS because it was just too hard to get all the dates rates in. I'd not focus at all anymore on building a super fancy customer front end because as said customers continue to write emails etc. So um there's probably a path on my A was when let's be much more a freight forwarder you know much more like sometimes cowboy style size of the company was more cowboy than DHL you know um and um let's

cowboy meaning aggressive aggressive in pricing aggressive in price like everyone every operator owns the GP of the file you know and not be too academic about all of this but every operator is an ATM owns the GP of the file um when someone comes and you have don't have something in the portfolio well you build that as the customer wants Yeah. Um and so on. The other option would have been and I think sometimes we were a bit too polished. We had a bit of a too polished, too much tech company as in let's be super focused. Everything needs to be great and well set up and we need to be perfectly fair to customers and so on. Um I think we had too much of that perspective sometimes. On the other side, I think the path could have been to your point, let's build a CargoWise competitor. Well, 2015 I remember was probably the year where DHL announced a rather big write off um and tried to do the Accenture and the the was the IBM SAP uh

Yeah. Yeah. Exactly. Yeah. And I think it it didn't work or it was too big of a of a project or the priorities.

I was part of that team. Yeah.

Oh, you were part of that. Yeah. So I actually I'd love to talk with you about that later but um should be should be doing a podcast about that actually because I think it's an interesting lesson especially today's perspective. Everyone thinks again about building CargoWise themselves. It's a great idea actually to bring actually I could bring definitely the leader of uh the one that was leading. I was part of some of the workstream. I mean was divided by workstream. There was commercial workstream. I was on the commercial workstream for pricing and rates and all this but they were it was a big project.

Yeah.

So I think and already in that year probably CargoWise wasn't a super modern piece of technology anymore. Um but I think also at the same time cargo wise um is a really powerful piece of technology because you know it's integrated super well around the world. It's got sort of I think the customs integration. So most of the important regions you can do almost everything with it. Um and so I'm debating with myself whether um you know building a TMS pre the AI era would have cut it. Um why? Because I also feel that um you know if your current TMS is a ladder not too bad but Russian car. Um and um I'm a new TMS

L I don't think people will uh will will

Okay. Um like it's a crap car. Okay. So, your current TMS is a broken car with three wheels. It's a broken car with three wheels. And um it's a broken car with three wheels. And I come and I'm a new TMS provider. I'm a startup tech TMS provider. And I tell you, hey, Mr. CEO of like, I don't know, 2,000 ops person forwarding company. I have a new TMS to offer is a Porsche and is like perfect, you know, perfect UX and also quite well integrated and so on and is almost for free. I don't think anyone would have I mean, I don't think at scale people would have bought it. Why? because the CEO says, "Last time someone promised me a new TMS. The project was like 3x more expensive, took 10 years longer. Um, and my people still work with emails. Um, and I didn't see the productivity went up like crazy." So, I think, you know, yeah, you could have probably built a better TMS than CargoWise, but distributing it into the market, I think, would have been insanely hard.

Yeah. Yeah. Yeah. And it's true also that 2015 it's important to know that at the time CargoWise was yet still under the radar right I mean the DHL announcement which was in 2016

that was what created the 25 billion valuation

gave gave a huge momentum because at that time DHL was also the largest forwarder yes at that time

and now they're going to be again.

Yeah I mean yes so I think I think that was very uh yeah very important going going back to Zaubert. So you um you talked about this 80% replacement um of the workflow. Does this means we replace 80 we replace 80% of the workforce?

Um I think so I'm not at all seeing that. Plus I think there's many studies that say we're not seeing that. Um but I think what and so the answer is I don't know. Um and I would think that however definitely the roles are going to change. I think what is very very clear today is that you will not anymore have to type 20 line items into let's say CargoWise or whatever um and then I will take care of that and I think you will more move to observing the process um jumping into exceptions um consulting the customer probably you'll do maybe more intensive QBRs you can do them more often you can do them also more informed you can be more sort of I think the strategic sort of almost um consultant um and move away from all the operational things. Um you have more time for phone calls when something when [ __ ] hits the fan and so on. Um so I think that's the roles will definitely change. Um I don't at the moment at all observe that um there is less people in in freight forwarding. Um but it it could be the case in 10 to 15 years. Yeah.

You think it's that far away?

It could.

I have the I have the feeling that it will be faster.

It could be much faster. Um but as said I

maybe not at 80% level for sure but but uh but that we would see a reduction of the workforce um in freight forwarding particularly because it's still like human capital intensive the work

the question is what is faster when you work on a freight forwarder today and I see that for all of our customers around the world many say I have real problems finding great operators um that are really knowledgeable that are always on top of their customers needs. Um and uh in addition to that uh what people in the US tell me is not the most sexy work category to get in and become an operator. freight forwarder clerk for example. Yeah.

Um in Germany at least for example you have the apprentice system where you have like great people that turn out of that but in many other parts of the world you don't have that. So

yeah I mean it's interesting because like in Europe in Germany also in France supply chain is a career. Yeah. In many places it's not at all. is a is um and uh I remember yeah anyway so um and and so I think AI is an opportunity for filling that gap of finding really great people. I think in addition what we're also seeing is um the supply chain workforce mostly in the U North America and in Europe is rather old and around 10 to 15% of your people um might churn anyway because you know they go into pen um because they maybe go to other companies as well etc. So the question for me is also yeah sure there will be some form of sort of time switch from a human to an AI but I think the 10 15% of your anyway turn in your organization might outgrow um the pace of adoption of AI.

Yeah. Yeah. Yeah. I mean there is obviously the opportunity also to serve a larger business meaning grow without hiring or not replacing a the workforce as they go into retirement or whatever. But but I do think we're gonna see particularly in freight forwarding or in some of the of the job categories that are heavy in terms of like what you said data entry processing and so we're going to see an important

to my earlier point about the CEO having a strong understanding of the what what it means for their company in the strong narrative.

The biggest problem I think with AI is thinking that is a zero this a zero sum game and that you know everything is dystopian and a problem. Um I think I advise everyone to try to say the cake is going to get larger and so if we have more time to be there for our customers let's try to do some of the things that are not so super commoditized. You know I can do I can move more into some value-added activities. I can move more into like purchase order tracking. Um I can make sure that I touch base with the customer often. I can move into other value added services. So there's so much opportunity that the world will also learn to create again the future um that I think time is now free for.

Yeah. I mean there's so many so many areas of opportunities, right? And and you on the operation side also like you know at for I think when you're just on the tech side maybe you don't see to that level but when you operate you see

the amount of opportunities that are

let's compare one more quote. Let's you know go one step further to make the service customer well. How many how many often days do you have not have time to just you know do the final 10% to deliver great customer service. So I think it's um yeah

great um Eric I think we could talk for hours but um just um let me ask you two more questions. One is what did you learn in Forto that is the most impactful for this new gig

from a company building perspective? So obviously I've done 10 years threat forwarding and I would say now I'm I' know more about threat forwarding than 10 years ago. Um but uh and that's a hu super big um advantage. It's a big advantage because you're you're a tech person that has that has run a freight for

I can go to my customers and I can speak at I think I can speak mostly at eye level not I mean a little bit below I guess but um and say hey look I understand your business understand what you concerned about every day and also in our deployment team for example was very important to me that ex freight forwarders work on our deployment team so they can speak eye to eye with the people in operations that were going to use our product and not some tech aliens from I don't know San Francisco or London fly in and like tell the operators to use AI now anyway so with regards to company the end is I think how people are perceived you know when they

do

some FDE it's it's the cool uh the cool job

hottest job in San Francisco that I think is also a very good job at um so about building the company and learning from Forto I think one of the things that is super important for me in this new business is um that every person is technical um and there's or has been vibe coding in the past um and has a like has that intuition and is like above the holy [ __ ] app um and understand sort of what's happening. Um it's just not possible to I think deploy such and work with such a product if you're not if you're not very technical. You need that intuition that's built from being exposed and trying it every day. So even the person who like Christopher who's in our deployments team is an ex McKinsey guy. Um uh he he um I think he I recently did the math I think he does 1.2 pull requests every week. Yeah. Um so pull request is a is a sort of a midsized for example code change to our code base. Yeah. And I think that's really great because um it's important that everyone who works with our product is deeply technical. The other thing that's really important to me is that everyone on the like very technical team and the engineering team is very very um focused on customers. Um because the other thing is also true. It's very hard for someone who's a software engineer to understand how threating works. And that is one of the key learnings from Forto. Um I had a really great team at Forto. But one of the things that we that I think we did wrong is engineers was so far away from the business that maybe after a year it was still not clear what the difference between a house and a master bill of lading is and I think that you cannot build a great product then um so I'm trying to make sure that both things are reintegrated engineers um I think all of our engineers have two to three customer calls every week and on the other side um everyone who's not on the tech team um is very technical as I think that's super important because you need to fuse now both worlds in order to be really successful.

Yeah.

Um so that's one of the things that's really important to me. Um and I think one of the things that stands out most from Forto.

Excellent. Yeah. I think that knowledge of the industry combined with technology is is definitely um going to be very valuable for discussions with customers also to understand be able to develop the right things. Yeah.

Um

who do you look up to? who is um you know who is inspiring you know from think about like founders or executives or would you look at cup to a

industry or or or beyond? Um, a couple of people I guess um, a couple of people. Um, I mean people that definitely inspire me are um, I think sometimes for example um, that board member that I was talking about of our customer who is I think you know running this huge organization and using Claude Code. Um,

not the one from Shanghai to Hamburg. not the one from no no one of the one of our customers not the board member of Forto um I think you know I do really admire how he's able to do both run an existing incumbent traditional organization and do that seemingly really well and on the other side I think you know be I I honestly I think you could definitely out compete many of the AI founders with regards to vision about hey the future is the AI has a complete ontology perfect sort of digital twin perspective and then we need agents that do this and that and so on it's it's crazy I mean In a certain way, I sometimes feel that uh when we do a workshop with him, our engineers go on and say, "Okay, yeah, yeah, I now understand what we need to do." Um, so I do I think I do look up to people who are able to combine both of these worlds. Um, I do um who else do I look up to for a long time? Not really thought about that. Um, I I sort of um can't deny that um I mean I've been going back and forth on this personality, but I can't deny that some like listening to podcasts from Elon is really inspiring. On the way here on my on the on the long haul here, I listen to the recent Dwarkesh podcast about how he's thinking that data centers will be bu built in space at the moment. And what just like about like about him and I think it's so important these days is this first principles thinking um and like rid yourself of all the legacy of the emotions. Um and like go back to first principles thinking. I really um I really um like that. Um and I think something I that also I think is really uh an important asset um uh and someone who's doing that really well I think is is might sound a bit cheesy but my wife I think who's a um who is while she's not technical she understands AI really really well and um I think she understands really well what the dynamics in society might be and um how we as a society I think should think about reacting um to AI is coming and then crafting I think great stories from that. Um I think it's so important that I think the people who understand what's happening that maybe top 3% or 5% who have tried cloud code at the moment that they um sort of almost for the good of mankind try to um I think shape to a positive the future that the world is headed into right and I think so we're trying to do um part of that as a family a little bit by also being exposed to politics and politicians because I think it's so important that everyone understands sort of what will happen in the future and I think um you know if Um, if you're good if you're good at sort of I think um bridging the gap between the tech alien language and root grounding that in the real world I think um then that's a really important skill these days.

Yeah, of course. What's the name of your wife?

Judith.

Juliet.

Judith.

Judith.

Judith, we will have you in the podcast at some point.

You should absolutely have her in the podcast and um Yeah.

Yes. Excellent. Well, Eric, thank you so much for joining us. It was great conversation flying. So, where do people follow you? What's the best to reach out if they want to know more about the company?

Um, the company is called Zauber. Um, Eric with a K at zauber.com or LinkedIn. Um,

yeah, we'll put the links also and information the description. So, on LinkedIn or direct email. Thank you so much, Eric. Great to see you. Yeah,

to be here. Cheers. Bye.

Thank you very much.

Thank you.