# 97 - 200 new Shopfloor Apps in Six Months — A Conversation with Eden Chabanet and Johannes Mohren of Arctory

[Music] Good morning and welcome to Entrepreneurial Relities, the podcast of the venture lab of the unique technical university. My name is Antoine Le Boirey and I'm the managing director of the software and AI, Tomb Venture Lab. My guests today are Edens Chabane and Jonas Moran, the two co-founders of Artorie, a Munich-based platform that lets manufacturing team build software application by describing them in plain language. Edens came to Artorie from Bosch, Exx out where she was at the heart of industry automation and Jonas before Artorie spend years implementing digital application in industrial companies. Between them, they carry the two world that Artorie knows sits in between Artorie software and factory floor. From a single request, the Artorie platform generates the operator interface, the database, schema, the workflow, the connector into our PMS and quality system. In the first six months, it has been used to create roughly 200 distinct software applications and customers such as Venture all day, I hope I'm pronouncing the name, okay? This is the platform which is used for daily shop floor management, ball, the escalation, analytics dash balls. The founder described Artorie as lovable for the factory, but govern, secure and built from an environment where probably the concern is not good enough. This conversation can use a thread we have had across several episodes which is below the rise of vibe coding. And the case for focused AI platform of our general purpose tool, Artorie is at the intersection of Bosch, Edens and Jonas, welcome to Artorie. Thank you for having us, Andreas. And then let's start with you. What in your experience convinced you that shop floor need at its own AI platform and not a general purpose tool? In my opinion, shop floor knows exactly their processes. They know what they need to build. And the tools that are out there need to fit these processes and not the other way around. And therefore we need to empower them with the right technology that understands their needs and builds exactly the applications that fit their processes in a daily, in a daily life. And that you cannot achieve with a general purpose tool because at the end of the day you have to do these compromises on your very perfect process to try to fit it in a box that was meant to do everything and at the end of the day it does nothing. What about you? How did you come to Artorie? And I know that you have a new -sual experience. Yes, basically as you mentioned, I spent a lot of time of my career on the shop floor developing these digital applications and learn that if you go very close to the process and listen to the people, you can create something very powerful and impactful. But always the challenge for me was that the backlog was over 200 ideas, big salts. Very difficult to basically manage this with a very, very small team. And in the other hand, last couple of months or like around one year ago, I witnessed that a lot of these ad-lib providers majorly focusing on the software benchmarks. And in AI, whatever you are measuring against is going to improve. Basically, they are very conviction that come from holding is going to be the number one case that is going to take off and basically you didn't see, oh, well, what I'm doing every day, ideally can today and can become doable by agents. From your own personal experience education, do you come from the world of coding? Do you own from the world of computer science? How do you come from the world of automation? We did it first and you're nice after. So I studied industrial and robotics engineering and that's where I started my career by deploying physical robots. So the first robots, the first AMRs, but as soon as you're deploying high tech like that, you always need to write a piece of software, a piece of code, a piece of program. And in the last 10 years, that's what I've been doing with my background in manufacturing and machines. I build software for manufacturing over and over again. So that's the story behind after it. For myself, I basically started coding at the age of 12 and basically did fall in love with developing software. But I decided again, studying computer science because I had a feeling I'm already able to write software. I don't need to study it and decided to study mechanical engineering because I wanted to find my industry where to apply these skills to me. And in mechanical engineering, I did find production technology. And so I did find there was a very, very big need on the shop floor for people that bring in those skills, but are able to bridge these two worlds together. So you need to know your way around the factory and the shop floor to talk and understand these people. But also it's good to combine with this skill because there's such a highly demand for people with software capabilities or digital capabilities inside the factory. Was artorees is now your first idea or did you have other elements then you pivoted? We had lived the pain of building software for manufacturing and especially digitalization applications over and over again in the last decade. And from that and with the capabilities that the LLMs today are offering to increase the velocity with which software is written and built. We came to the conclusion that the biggest problem in manufacturing that nobody was able to solve, which is scaling digitalization is now going to be solvable. And we are having this unfair advantage and the right positioning to build this next platform that scales digitalization manufacturing. So we were spot on from the beginning on the idea and now we're finding more and more customers that are going viral with this solution. Yeah, it's basically that's no question about that there's a huge problem in demand. And also our vision was okay. We want to empower these people. And this didn't change over time but of course finding the right product, finding the right language. I mean that's still obviously evolving in artoree but also inside the industry as well. Can you describe what happens when you use the platform? Can you describe in practical element what takes place? Yeah, so basically the vision is that a person who's not a software developer can just describe their idea without having to think about the tech. So I don't want the person to describe or need to connect to an API. I need to implement like a hook or the front-end component. I want them to think about the process and about the end use of this application. And then just describe a natural language to the AI agents what they're thinking about. Basically after they send out the prompt, our agents are very well and very known to our framework and template in our platform. They basically build these applications for those people. So as soon as you send out the prompt, the I first gets familiar with the framework and then basically explore some use case examples we have in the platform and then it starts implementing your idea from the front-end, but also the database aside. So if I understand well, you got 200 applications which have been deployed with the platform. Do you know if they are regularly used, maybe you've got over 200 since we last spoke? Let us go back to the earlier question. I would like to actually give the point of view of the user. In the before world they had the problem and they knew how to solve it from the shop floor. And the workflow was they either went to the IT that they said, "Okay, we had no time to implement this for you or maybe in three months or they had to build it themselves in tools like Excel or other tools that are not for software development. This is a before world." In the archery world or the archery world, the user goes on the platform, knows exactly how to describe their problem and the outcome that they want to have. They write it in a few sentences and then a few minutes later they have the full stack application that is ready to be deployed for their own use in the whole team and their problem is solved. There is no compromises, what IT is able to do and when, but there is this empowerment and pride of "I solved my problem and all of my team can use it and now we can work better together." Sorry for breaking your flow, but I felt this emotional point was really important to bring through. Okay, so do you have 200 applications or 300 or 100 how many? I mean we are very aware of what these applications are and how they are being used because with every customer of ours we have a office hour once a week. So we sit with them together, they can come and oftentimes they just come because they are so proud of what they have built and want just to show it to us but they can also obviously come with their have like a challenge they are not able to do. this and it helps us helping obviously the customer but also learning where our platform is failing at the moment and basically be able to improve that the next person doesn't stumble across this problem again but therefore yes we know these 200 or above 200 applications very very well because we're sitting with these people and they're very very in love with them because they have built themselves they're very proud of that and this also helps a lot in the adoption side of the shop floor so it doesn't come from somebody external or top down as the people that have built by themselves and this is also why they are very very optimistic in adopting these solutions. I think that you use the comment when we discuss you said that vibe coding is addictive so is vibe coding for shop floor also addictive? Yes I mean of course I am basically our convictionist that people inside factories they're billers by heart I mean their entire job is to build products every day I mean this is what the factories about and so far just from the digitalization side they didn't have the my tool on their hand on being able to shift this mindset into the digital world this is why for us manufacturing this D number one place for vibe coding because the people already have the right mindset they are to build something. Then you give specific examples of the type of mitigation which I've been I put it I mean you've described the the the the process you've described the emotion about having this application. Can you give examples of software applications which aren't typically deployed with your platform? I mean one application we start very often with is the shop floor management so maybe for somebody that doesn't come from manufacturing it's the equivalent to the stand-up for software teams so every day they basically get together and look at their most important KPIs oftentimes this contain safety so was there like an incident yesterday they look at quality they look at delivery costs and people and then basically from this they divide some corrective measures and beforehand the majority of them tractors on paper which obviously makes it very very difficult if you have over 100 over 200 different corrective measures to still get the overview of what has been done with responsible what does the due date and this is obviously where software can help a lot already and then from there the oftentimes go into different functions for maintenance another very important use case is they're printing out their production orders from the ERP system and then hand them into the production which even oftentimes leads to their these pieces of paper gets lost and then there's no traceability so oftentimes a customer question says hey where's my order and then you basically try to trace down the paper and just basically find out that this paper did for you on the floor and then was picked up by the cleaning person and then was thrown into the trash and suddenly the customer order wasn't produced anymore so they think many many many of these applications which maybe are now becoming too long to answer the questions but there are like a lot of ideas a lot of ideas at the moment living in the backlog of these production teams that basically just waiting to be digitalized what is your typical customer we focus on the machinery and equipment manufacturers that are mostly island based so we're not focusing on automotive where it's fully automated there's a lot of data and data plumbing need to be done and why we focus on these people because their backbone is mostly the ERP today but the solution to digitization and to cover all of these use cases you have us talked about is not another ERP module it's a completely new different layer that allows them to digitize and scale but gives them also the flexibility of evolving their system landscape where the ERP is going to become in the future just a database and a system of record and arteries going to be the place where everything is built all the intelligence are the interaction with the with the users they have a lot of work stations around so that a lot of humans and workers on each machine and these are the people we are focusing on and we are starting with with Germany do you have also interesting other countries are you just right now in Germany? I mean yes we are interested in scaling in Europe so we already have ideas about going with customers in France and the Dach region etc but the initial challenge for a startup is to focus focus on one ICP on one region to deliver and to the hidden champions in our backyard here sitting in Munich and then go in European and World War. Janis one of your thesis is that top-down digitalization failed because it is detached from the process and you've also called layout because customization hidden structural that can you comment on these things? Yeah, it's basically inside the factory even before the utilization of this entire lean movement there's always the idea of solving problems for nobody so you want to empower your people to take actually you don't want if they have like a small problem to try to escalate it to the very top manager we want them to solve it where the problem is because they know the best the problem they know the best solution but at the moment this doesn't transfer into the digital world so if somebody has an idea the normal way of as he described it earlier is that they need to go to the AT and basically ask them hey do you have time to build this for me? At the moment all of them updating the ERP this is very come to the technical depth so no they don't have time with this why for this right now the other way is okay maybe I can buy extraneous solution but this will not fit their process and also they stumble across the question do we have budget is here? so this is why a lot of this idea gets stuck in the back log and do not see the light of the day and what we want to combine is this already existing mindset in the lean of solving the problems locally but in a digital world how do you position what you do with respect to either the ERP players that you've just mentioned but also the the generalize vibe coding platform like clod code or loveable or you know other guys or codex yeah I mean very good question so maybe starting with the ERP so I mean at the moment go number one isn't to be placed the ERP from day one because it is not that we go into a factory and it has already been fully digitized with the ERP the biggest problem is they're like so many processors that have not been digitized yet so this is definitely where we are starting in the beginning that basically all of these ideas that got stuck in the back log we want to bring into the real world and therefore of course we are sitting the bath or sometimes our customers say about the ERP system because for a lot of applications you need to get data from the ERP and bring it into our application in our to me and comparing this exactly against solutions that are on the market so yes there are a lot of coding solutions but I always say basically are to be dissolving the last mile of software creation because yes everybody can take a software development tool or like a lm and basically write code but then you quickly get into okay now how do I deploy this and basically double build started with the idea make it easy to deploy but then very quickly you get from basically what are onto the nervous to do or what somebody who does their side project into enterprise applications and then you get questions like okay how's the user management handle how do you do world-based access management how do you do auditing how do you do versioning how do you do world-back so this is all you need to do when you look at enterprise application and after we put this into one platform what about the IT department it may be maybe I see it in left things so it and for you let's put ourselves in the shoes of the IT department they want something to govern to have an overview to actually avoid shadow IT and something they feel safe rolling out to the entire organization and this is exactly the anxiety that we usually are facing with these IT tools and platforms and that's what archery also offers by definition and by default in the platform as Johannes just said so yes it's an anxiety IT can think they hate us in the beginning but then when we show them the platform that they do not need to do the heavy lifting and they will bring everyone in one platform to build in coherent cohesive way applications that respect the policies of the group that look the same and they are secured by design then they are actually excited about archery there's another type of player who is often very active in the environment which is the local systematic writer so what do you think is going to be happening with Clio? classic integrator when we are told that we are going to be basically being used. Classic integrators are a couple of possibilities in there. You have the integrator of machines today that are actually lacking this ID and AI capabilities because basically these are talents and resources that are very very scarce in manufacturing but their end users are asking them to deliver machines that are connected to speak the same language, these KPAs etc. So there I think the resources are lacking and for the software integrators let's really look at have a hard look at the market who is able to afford a software IT integrator. It's only the big companies that are already in automotive and big aerospace and I have not seen a middle stand being able to recruit one of these IT software integrators. They cost a lot of money and it last also for two three years and it's easier to start it in a green field factory rather than brownfield. Have you seen some of the latest stage in technology being developed by big incumbents like SAP and how do you compare what they do with what you guys do or what your user are doing? I mean maybe looking at SAP I think SAP has a very big problem with companies that are already in their systems and everybody knows that who has been working with SAP how do I find the right data and basically before an SAP you didn't need to know the transaction, the variants basically what filters are applied and this knowledge it used or lived at the moment in the hands of these key users these are what they are called so even when I was working with companies I wasn't able to navigate the SAP by myself even though I'm a digital person because all of this knowledge is unavailable. I think this is the first power for SAP and what they're focusing on right now with L&M's how do I make this data easier accessible so basically talking to my data and can you please show me the orders that are going to be laid today before you would need to have been a key user in order to find out this data and then do it in Excel. So I guess this is what they're focusing on first and then the second idea of what they're doing now is okay how do cannot I automate this work being done in SAP by users at the moment with the iH list and I think yes this is very valuable but then on the other hand especially when we look at manufacturing we have such high standard of quality that always these stochastic and probabilistic systems struggle in decent environments and it's less of a technical hurdle it's more of a cultural hurdle and oftentimes say I've like two hearts in in me is why I'm the production engineer and I'm the data scientist and the data scientist says 99% accuracy amazing I've done a really good job and then the production engineer in me says okay 99% accuracy I would never use this in manufacturing because there you have this mindset of six sigma so basically if you're for example manufacturing car brakes you cannot say one in the 100 car brakes who cares about it it's a good quality and so this clashes culturally and I think they are for this notion trying to use LMS at the moment to do this entire work is valuable in some areas but I think we'll have a hard time at the moment the technical perspective to be adopted on the short fluid set. Then you mentioned the word viral which earlier you told me is not a word that is usually associated with the industrial word so how long are you guys growing? Let's look at the last eight months and the number of applications actually multiplied by 34 times since since we actually launched and that resulted in the couple hundreds that you have already today what we see is that we start with one person one team and in two weeks we are already moving from just the shop floor team to the quality to the maintenance team and I think in the very first week we launched we went from one key user to 30 users by the end of the week so in manufacturing we have never seen something go viral this this much and there's also this misconception that manufacturing is one of the latest places to adopt software and technology yes because software was never built for manufacturing needs but this changes today and and we see it in how how proud the people that build the applications are and how many users grow week after week since lunch so so I mean this is the growth we see inside companies but at the moment we also see that actually our biggest sales channel are our existing customers because they're so proud of what they're doing and they're so amazed by us that they invite the local network because they're already in these local networks to come to their plant and to basically just show them the factory we show what they have done with the after we and then we basically take it over and do a small hackathon with them so we give them access to activity they're allowed to build the application they can take it home and then obviously this is this addictive part then they're hooked now I want to use it because then now see what is possible and then you go back to your old world where you have to query the data here, query the data there but then excel make a new IT ticket because you want to have this then it gets into a backlog and you always every day think about I could have already built this an up to me and this obviously leads to that these people then want to onboard activity and then the effect takes place that Eden just described once you're in the company very very quickly within a couple of weeks and months basically they have already built a lot of applications. In our last hackathon and four hours the customer team built over seven different applications and one of them was built in four hours while the person who did build it said the exact same replica that I showed you this morning took me six months to build with the IT. Wonderful let me ask you the classic questions do you have a new advice for our listeners? This is a very good question I think my advice to the listeners is to never really accept the status quo the cliches about the market or what was in the past possible or not possible everything is basically changing today and also to really listen to the customers listen with a big empathy they tell you what they need because transformation is only 20% technology and 80% people and our customers even though we work in B2B they are people at the end of the day and maybe to add to this and maybe to give it a name look for the controversial tools so what is true in the industry because it's broken but not a lot of people today agree with you on me I think this is what Eden and me combine in our past experiences that a lot of times we needed to find the status quo and it also was the case with our to be where we started last year everybody thought we were a little bit crazy like why do you want to do software development and manufacturing and now the shift that is here that suddenly everybody say I of course it makes sense because it solves all of these decade all problems but in the beginning nobody believed us until we did show it I think the advice for the listeners is go into market and look for this and I think this is especially advantage for a young person because you are naive and you are allowed to be naive and you are allowed to question the status quo. Thank you very much. Thank you very much. Automatically it is available on major podcast platform when you can find other inspiring presentations do subscribe if you love these podcasts and want to hear more do give us a rating let your friend know about it and we look forward to having you for more entrepreneurial meetings