Central Service Platform: Murrelektronik & elunic digitize product data

Today, we’re diving into a topic that’s top of mind for many in our community: How can digital services in industrial environments actually be brought to life – despite data silos, complex system landscapes, and internal challenges?
Joining us today is Murrelektronik, a recognized specialist in decentralized automation technology. Many of you might know them from their work in control cabinet environments, I/O systems, or industrial connectivity. In this episode, they share how they are strategically positioning themselves in the IoT space.
Also with us is their technology partner elunic – the team behind the modular platform solution with integrated AI that we’ll be exploring in this conversation.
You’ll get hands-on insights and best practices, including how decentralized product data can be transformed into a centralized digital service, why the Asset Administration Shell serves as a structural foundation, what this looks like in day-to-day operations, and how the AI assistant ShopfloorGPT is already answering real-world questions from machine builders – even without perfect data, but with real value for service and sales teams.
Joining us are Fabian Distel, responsible for digital transformation at Murrelektronik and Managing Director of Murr Shared Services GmbH, and Jonas Schaub, Executive Board Member at elunic.
As always, you’ll find all the details at www.iotusecase.com and in the show notes.
With that – let’s go. Enjoy the episode!

Welcome, Jonas and Fabian – great to have you both with us today!
Fabian, I’ll start with you: How are you doing today, and where are you joining us from?

Fabian

First off, thank you so much for having me today, Madeleine – thanks for the invitation! I’m doing great. I’m currently at our office, in between lots of meetings as usual, and I’m really looking forward to spending the next half hour with you and Jonas.

When you say “office” – where exactly are you based? Can we give a shout-out to some of your colleagues?

Fabian

We’re spread out across Germany and have many offices. I’m currently at our headquarters in Oppenweiler, Baden-Württemberg – that’s near Winnenden and Backnang.

Then greetings to the region – and of course to everyone tuning in today!
Jonas, a warm welcome to you as well! Great to have you back. I think you were actually one of my first podcast guests, which makes it even more special to have you here again. It’s an honor. How are you doing today?

Jonas

Thank you, Madeleine. I’m really happy to be here again and looking forward to our conversation today.
Great to hear that we kicked things off back then. That must’ve been five years ago now, right? Back then, the world was a different place – lockdowns, COVID – a very unique time.
A lot has happened since then. I’m doing well, thank you, and I’m really excited about our conversation today.

I’m also really curious to hear what’s changed for you since then – which is why I’m especially excited about this episode.
Let’s start with a quick round of introductions. Fabian, I’ll begin with you:
You’re CDO at Murrelektronik and recently also became an authorized officer for your service unit – if I can put it that way. I believe you previously worked at Windmöller & Hölscher, focusing on digital services.
And if I’m not mistaken, you studied at TU Munich and were part of the management team at IWB, the Institute for Machine Tools and Industrial Management, where you also did research. I hope I’ve got that right!
Great to have you on the show today!
When you look back on your experience – was there a personal “aha” moment in the development of digital services when you realized why they actually work in the industrial world? Would you like to share something to kick things off?

Fabian

With pleasure. Just to clarify: I’m an authorized officer at Murrelektronik and Managing Director of Murr Shared Services GmbH.
I never quite knew what was coming next – and that’s exactly what makes it exciting. That’s still what drives me every day.
Whenever I talk with Jonas, we instantly come up with five new ideas, and suddenly the next six months take a completely new direction.
That’s what makes it so much fun. Especially in the digital space, there are so many opportunities you can seize. In the end, it’s always about setting priorities – just like in life – and then getting started.

That’s a great mindset. And you have to be someone who really drives things forward, especially in a market that’s constantly changing.
There’s a lot happening – it’s a dynamic environment, but not always an easy one.

Fabian

Absolutely. But you need to enjoy the new, have a bit of an adventurous spirit, and try to look at things from a holistic and systemic perspective. Then it becomes a great opportunity.

Yes, totally.
Before I turn to you, Jonas, let’s briefly talk about Murrelektronik:
Most listeners probably know the company. Your core business is in decentralized automation technology – power distribution, communication interfaces, power supplies, and so on.
What’s currently happening at your company when it comes to IoT?

Fabian

Sure. We come from a traditional component business. But what has always set us apart is that we try to create smart solutions from the customer’s perspective. That means we look at challenges through the customer’s lens.
In electrical installation, the typical questions are: How do I plug something in? How do I connect components efficiently?
But when you take it one level higher and look at the customer’s entire operation holistically – including their processes – you see that a digital transformation is happening. Everything is being digitalized.
This means: Today, we don’t just have to provide smart hardware solutions.
We also need to think about how to make working with Murrelektronik as seamless as possible across all business processes.
And that’s the exciting challenge we’re tackling.

Maybe one quick follow-up question:
What’s a typical example of a component that you produce in your core business?

Fabian

One example would be a cable with two connectors – typically M8 or M12 – standardized plug connectors that transmit analog and digital signals as well as power.
We also offer I/O modules, meaning switches for field use, decentralized power supplies, and a wide range of electrical components in general.

That’s a great mindset.
Before we get to your customers and the specific challenges they face, let’s briefly shift the focus to you, Jonas.
For those who don’t know you yet: you’re a board member at elunic AG, and as you mentioned, you’ve been a guest on the podcast before. Back then, we talked about customer portals in mechanical engineering.
Today, we’re picking up on that topic again – in a more advanced and evolved form, and in a new context. You’re currently working with Murrelektronik on a very concrete project together.
What I’d really like to know: In your view, what has changed on the IoT market in recent years, especially in terms of digital services and related topics? How do you see the current developments?

Jonas

On one hand, generative AI has introduced a completely new technological dimension. Ideally, it builds on certain preparations and enables capabilities that simply weren’t possible before.
Looking back at the last five years of IoT, I’d say my feelings are a bit mixed.
There was a lot of hype – along with high expectations – but also some disillusionment when it became clear that not everything worked as imagined right away.
I often compare it to the 1990s: Back then, the business model of the internet was just to get people online – remember Boris Becker and the AOL CDs some might still recall?
It’s the same with IoT today: The basic infrastructure – the networking of things – still isn’t fully in place everywhere. Connectivity is the foundation, and real value creation builds on top of that.
What we’re seeing: Many manufacturers are struggling to productize their solutions – meaning turning them into real digital products that operators and producers can actually use.
At the same time, when we look directly into our customers’ production halls, we see impressive solutions emerging – from predictive maintenance to smart networking of systems, components, and machines.
These enable massive efficiency gains.
A lot has happened over the past five years, we’ve seen a great deal, and now even more possibilities are opening up – if you approach things the right way.

[08:39] Challenges, potentials and status quo – This is what the use case looks like in practice

You’re sitting together here today – can you walk us through what exactly you’ve implemented in this project?
And how did it all begin? Is there a story behind how your companies first connected?

Jonas

Maybe just a quick note on elunic to start: We’re based in Munich and have three main business areas.
We ourselves operate and manufacture inspection systems with AI – AI.SEE™. We introduced these briefly in a previous podcast around five years ago.
We use our own technology to maintain those systems efficiently remotely – meaning fewer technician deployments and reduced travel time. Maintenance and other services are therefore carried out remotely.
The second business area – the one we’re talking about today – focuses on supporting machine and component manufacturers, like Murrelektronik, in building digital solutions.
We rely on proven technological building blocks. The topic of the “customer portal” that you just mentioned has been with us for some time and is now being continuously developed further.
This isn’t just about incremental improvements, but also real leaps – for example, through new technological possibilities like generative AI.
I believe this very aspect – the customer portal and the empowerment of Murrelektronik’s users to access services themselves – was a key starting point in this project.

Fabian

Our collaboration goes back to my time at Windmöller & Hölscher.
Even back then, we had very productive conversations that I remember fondly.
When the task at Murrelektronik came up to build a comprehensive foundation for our digital customer offerings, we quickly reconnected.

Let’s now dive a bit deeper into your project.
To begin with, perhaps a structural question:
There is the Murr Shared Services GmbH and, in addition, the Murrelektronik GmbH.
Fabian, could you briefly explain how your digital setup is structured?
Are there different units? And what exactly does that include?

Fabian

With pleasure. Murr Shared Services GmbH is primarily the internal service provider within our corporate group.
Murrelektronik is not just a single company – we have over 25 subsidiaries worldwide, as well as several other entities that belong to the group.
For quite some time now, we have been managing the entire IT for all companies from the parent organization.
Since I joined Murrelektronik – which was a little over a year and a half ago – we’ve brought together various internal departments.
Previously, each department had its own digitalization unit: in sales, development, procurement.
These units were consolidated under my leadership and have since been spun off into Murr Shared Services GmbH.
From there, our goal is to establish a group-wide standard for digitalization. That’s essential, because successful digital transformation requires well-thought-out, standardized processes.
Only with such a foundation can we create modern, future-ready solutions, both in terms of processes and systems.
And that is precisely the role of Murr Shared Services GmbH.
To clarify my second role: Together with my colleague Dr. Paul Zeller, I’m building a new business unit within Murrelektronik called the Global Business Unit Digital Solutions.
This unit develops new digital products for our customers – in other words, marketable solutions, mostly in the form of SaaS products. Our first product is uKonn-X, which we are currently launching on the market.
So, you can think of the roles as follows: On one hand, we develop concrete, market-ready digital products. On the other, through Shared Services, we handle internal digital standards and services – for example, the customer portal that allows all our international subsidiaries to interact with their customers.
And today, this customer portal is much more than a classic portal;
It brings together many digital services, integrates various systems, and naturally includes AI components as well.

So just to clarify – this isn’t about your production or anything like that, but really about the digital service portal and the smart products you’re making available to your customers.
Did I understand that correctly?

Fabian

That’s right – but that’s exactly what makes my role so exciting: I’m responsible for both. From my perspective, our production is absolutely tied to the goal of providing data to our customers.
We develop, among other things, safety-critical products. That’s where traceability comes into play. We’re talking about Asset Administration Shells (AAS), which we want to provide to our customers in the future on an instance basis.
If you really think the digital thread all the way through, it affects all areas of our company – right up to the customer.

Before we dive deeper into the technologies, I’d like to touch briefly on the “why” – the underlying problem.
At the end of the day, a project like this is always an investment in technology – and not a small one.
And your customers are also investing.
So let’s talk briefly about what we might call the problem statement – or even the business case.
Why did you decide to invest and say, “This is where we’re putting our money – this is our future”?
What was the concrete trigger?

Fabian

We have many customer touchpoints.
If you look at the customer journey, there are different roles on the customer side – and different moments when we interact with them.
Traditionally, each of those touchpoints has had its own solution. And yet we’re often providing the same data – just in different formats.
Sometimes our sales team delivers the information, sometimes the customer retrieves it from our website or webshop.
It’s not consistent.
Especially from the customer’s point of view, it’s cumbersome – they always have to go to a different place.
Our challenge is to unify that information delivery and think of it as an end-to-end experience.
Ultimately, we also want to break down the digital wall between the customer’s system and ours.
We deliberately ask: What is stopping our customers from working efficiently with Murrelektronik?
And we want to remove those barriers, especially in the digital space.


There’s still a lot of untapped potential that we want to unlock.

Very interesting. You just mentioned that data isn’t consistent.
You have many systems in use – webshop, sales, various IT systems, and so on.
Where exactly would you say: “This is where we’re really losing time and money – and it’s unnecessary”?
Is it mainly about data management – a technological issue?
Because each system has a different interface, and a lot of processes are manual.
From your point of view, would that be the biggest factor where you said: “That’s where we need to invest”?

Fabian

Right now, we’re facing a lot of system discontinuities – both on the customer side and internally.
Because we have so many standalone solutions, data redundancy is almost inevitable.
That’s because data has to be prepared differently for each system – and that creates inconsistencies.
We’re tackling these challenges on several levels at once – but not everything all at once.
Instead, we break things down into manageable steps, where each one delivers a clear benefit for the customer, while also helping us clean up our internal processes.
The platform we use to enable integration with the customer is shopfloor.io by elunic.

Jonas, a question for you. When you hear all this – and considering you’ve handled many customer projects – are these the typical issues clients come to you with?
Do they say things like: “We want to use our data more consistently, centralize everything, and build something together”?
Are these the biggest technical challenges you see, or are there other key aspects?

Jonas

It’s actually quite a complex picture.
One of the central challenges is, as just described, centralizing and streamlining the customer interface – in other words, bringing all relevant information and features together in a single system or touchpoint.
Many companies today operate different systems – for support, spare parts, IoT telemetry, process data, and more.
That’s hard to manage – especially when there are multiple endpoints and login credentials involved.
One of our goals is therefore to consolidate all these topics in one central place – we often refer to this as a single interface or a single face to the customer.
a centralized entry point that makes the customer experience and usability much easier.
These portals can also evolve from smaller applications.
For example, someone might start by implementing a spare parts catalog – and then realize that it would make sense to integrate access to other tools as well.
This calls for an overarching portal structure that brings together different technologies under one roof.

You just brought up the topic of the Asset Administration Shell – could you explain that with a concrete example?

Let’s say I’m a customer and I buy a connector or an I/O module, maybe through your webshop. You want to achieve consistent data management.

Was the problem that product types or file names weren’t standardized?
Do you have a real-world example of that? Are we talking about product master data – and if so, what was the specific issue?

Fabian

Exactly, it’s mainly about product master data, which can change over time due to product updates or new versions.
Let’s say we adapt a product and extend the permissible temperature range by 10 °C. Now the question is: How do I document exactly when this change was made? In our webshop, the product simply shows the updated temperature range. But our current systems can’t version those changes or document them in a traceable way. And this is exactly where the Asset Administration Shell comes into play. It allows us to store such changes in a versioned format. Later on, when we work with instantiated AAS models, each individual product will be assigned to a specific version of that shell. That way, I can always say: This exact product was manufactured with the earlier, narrower temperature range.

Right. For everyone hearing about the Asset Administration Shell (AAS) for the first time – I think many are familiar with it by now, but just for context:
I’ve already recorded a few podcast episodes on the topic – I’ll link those in the show notes.
How do you define the AAS in your organization, or more specifically in this project?

Fabian

The Asset Administration Shell is essentially an API for product data.
There are different variants: You can download it as a file or container, or access it directly via an API.
The beauty of it is that it’s standardized and machine-readable, so it acts as a true interface standard between companies and systems.
That’s particularly valuable to us, because we want to build a holistic solution where product information is accessible across different services.
Instead of providing a proprietary data format for each individual service – including master data, images, or document links – we say: there is a single source of truth, and that is the Asset Administration Shell. All services access that central source.

Okay, and that’s when you decided to bring elunic on board to build this together?

Fabian

One of the many elements we’re integrating into our customer portal with elunic is the Asset Administration Shell server.
The goal is to use it in the future to manage both product master data and instance-specific data.
And everything we develop together with elunic—like the GPT assistant, for example—should pull its data from there, not from some separate system or isolated table.

And if you’re listening and thinking, “Hey, we’re facing very similar challenges”—I’ll include the contact details for Jonas and Fabian in the show notes. Feel free to reach out and connect.
Every project is unique, of course, but I think a lot of the underlying challenges will sound very familiar to many of you.

[21:40] Solutions, offerings and services – A look at the technologies used

So how did you actually implement all of this? What are the concrete solution components?
You already mentioned ShopfloorGPT; Jonas, maybe let me start by asking you: What exactly is this solution, and what components are involved?

Jonas

Our platform solution, shopfloor.io, is fundamentally based on a commercial open-source license.
This addresses several requirements—most importantly, it ensures that the company using the platform also receives ownership of the IP.
This allows us to avoid dependencies or lock-in effects—especially in value-creating areas.
The architecture is modular. We often refer to it as a “make-and-buy” decision—do I develop something myself, or use an existing solution?
shopfloor.io combines both: internal and external development teams can continue to build and expand the platform.
The platform is divided into various modules and services. A central core service is the asset structure— Nearly all additional services – whether product provisioning, support systems, spare parts catalogs, or IoT telemetry data – are directly tied to physical assets such as machines or hardware.
And all the other services are built on top of that structure.

Got it. Fabian, how are you using that in practice?
Are you deploying the platform more like a ready-made product, just rebranded under your own name?
Does the customer ultimately see your logo?

Fabian

We use it as a white-label solution. That means: green, with the Murrelektronik logo.
As Jonas already mentioned, we’re integrating various services into this platform.
A big part of this will revolve around assets, but also around the digital solutions our business unit develops and offers to customers.
We also use the platform for user management, including single sign-on for all digital products we provide to customers.
And we’ve got a lot more planned—many more services will be integrated.
And that’s exactly why I’m so excited about the ShopfloorGPT module.
Because we all know the problem: you have an intranet page with 50 tiles—and you’re desperately searching for the right one.
That’s what we want to avoid. We want to keep the services modular—for maintainability and scalability—but for the user, the entry point should be a GPT-based chat interface. There, they can simply type in their request—and land directly in the right service.
That’s really exciting.

Looking at the implementation, the very first step was probably getting access to the data in the first place.
So you probably started by connecting various IT systems within your organization.
I assume you’re not yet transmitting live data from devices you sell to customers—not typical IoT telemetry, right?
It’s more about product master data, like what’s stored in your sales systems, CRM, webshop, and so on?
Is that your current form of data collection?

Fabian

I’d say: We’re not doing that yet, but we’re actively thinking about it.
As I mentioned earlier, we’re working across many different dimensions.
Our approach is: Every step we take must create real value. That’s why we don’t implement everything at once—we do it in releases.
For the start, we focused on two concrete services: A product catalog and the provision of all relevant documents.
It’s not just about giving users a list of download links.
More importantly, these documents should also be made intelligently accessible via ShopfloorGPT.
Currently, we’re working in parallel on three additional services, which will seamlessly integrate with the existing ones.
The goal is for everything to eventually be accessible through our GPT interface – what we call Murrlin.

Jonas, a question on that: shopfloor.io serves as the technological foundation, and with ShopfloorGPT, you’ve developed an additional module.
How do you integrate the data into it? Is all of that already integrated into shopfloor.io, so that, for example, data from a CRM system can be connected directly?
How exactly does this work in your tool?

Jonas

Our modules are very advanced. shopfloor.io has been in development for around five years and is accordingly mature.
The platform offers a broad set of features that many of our customers are already using productively—even in end-user scenarios.
A good example is MyMurrelektronik, which is available online. I think it’s something you can explore quite interactively.
Let’s imagine a typical use case: someone wants to build a packaging machine for the food industry.
In the past, selecting products often meant having a sales conversation—which, from a process cost and efficiency perspective, wasn’t ideal.
Today, we enable self-service access via the portal: users can independently find the right products—filtered by product groups, technical parameters, and facets.
Thanks to large language models, it’s now possible to make this enormous product catalog—nearly 80,000 items in our case—accessible via natural language.
You can speak just as naturally as I am now. You no longer have to click through dropdowns and checkboxes.
That’s a major leap forward in terms of accessibility and self-enablement, especially for users looking for specific components.
What’s particularly helpful: language models don’t require a fully structured data base.
In traditional digitalization projects, you often had to harmonize and clean up data first— a lot of effort for little immediate benefit.
With language models, we can easily crawl and extract existing product information – for example, from the webshop – use it as a basis, and structure it.
This ties into the Asset Administration Shell, which provides a formal description of that structure.
But in combination with language models, we’ve now been able to integrate large product inventories without launching a costly data cleanup or migration project.

So, you’re using the Asset Administration Shell as the foundation to standardize the structures—especially the product master data?
And your AI-powered assistant then pulls exactly the product or configuration data it needs to process further?
Did I get that right?

Fabian

Exactly. From a technical standpoint, we work with a multi-agent AI architecture.
We have, for example, one agent that accesses our documents—the classic PDFs you can download for our products.
Another agent uses the Asset Administration Shell to specifically retrieve information from it.
That’s the current technical state.

[29:22] Transferability, scaling and next steps – Here’s how you can use this use case

If you—or others listening—are planning to start a similar project:
What advice would you give them?
Are there any best practices where you’d say, “This is something you definitely need to keep in mind when building an AI-powered assistant”?
Is the central topic really the Asset Administration Shell, or are there other factors to consider?

Jonas

I’d say there are several perspectives.
The first one isn’t even necessarily technological, because the whole topic of self-service and user enablement is often more of an organizational than a purely technical challenge.
I’m a strong advocate for it, because I believe in the enormous productivity gains that are possible—and I regularly see them in practice.
But: There’s also a risk of approaching the topic too technology-driven.
Based on our experience, we recommend starting with a clearly defined use case—one that works and doesn’t involve too many people right away.
That way, you can test the topic in a safe environment, without it becoming controversial or overloaded from the start.
If that first use case is successful, you can build on it. It becomes your first reference—a lighthouse that others can align with.
Another advantage in our case is: We already bring the product with us.
So you’re not starting with a theoretical concept or a mock-up—you have a working interactive foundation right away.
That makes it much easier to imagine a realistic target picture—and at the same time, you can quickly spot technical or organizational challenges.
The visual and interactive usability really helps here.

Very interesting.
Fabian, what does that look like from your practical experience—would you add anything?

Fabian

For us, it was extremely helpful not to start with the most complex technical solution right away—because that obviously takes time.
Instead, we began with a prototyping approach.
Before we built the agent structure, we simply loaded our data into a vector database and made it available to our colleagues in inside sales.
The feedback was surprisingly positive: “I’ve never found the information this easily before.”
Even minor inaccuracies in the responses weren’t a problem—our colleagues were able to handle them well.
That really helped us: We gained a better understanding of our data sources, identified gaps, and optimized in parallel—for example, by adding missing or unclear documents, or removing irrelevant ones.

Fascinating how you approached this.
I think with projects like this, change management really shouldn’t be underestimated.
Demonstrating the value first – in a small, clearly defined environment – is definitely the right way to go.
Especially with your example from sales: start with a specific use case, involve initial users, gather feedback, and then roll it out step by step.
I imagine that communicating such a project internally and getting buy-in must be quite challenging.

Jonas

Absolutely. And maybe one more thought to add: In the past, innovations often came from the defense industry, then moved into the B2B space, and only later reached the end customer.
Today, it’s quite the opposite. Many innovations are born in the B2C world – our private lives – and only then make their way into B2B.
A good example is the use of drones.
These developments are actually really helpful for change management.
They inspire people and raise acceptance for new solutions.
Just think about how naturally we use online banking or buy train tickets via smartphone – no one would seriously go to a branch or ticket counter anymore.
Technologically, this has been possible for a long time. And it’s only logical to gradually remove the intermediary – the human in the middle.
This process is called disintermediation.
And that’s a powerful vision, even in an industrial context.
Of course, it’s a transformation, and yes – it’s challenging, especially for people who are used to doing things a different way.
But this isn’t just a technical shift – it’s a mental one, too. That’s what it takes to implement this kind of change successfully.

Very well said! Looking ahead – what’s on the horizon for the next three to five years?
Jonas, earlier you mentioned that AI is now able to access data that may not have been integrated before.
Would you say that’s one of the key trends – AI enabling new value in IT architectures?
Or where do you personally see the biggest developments?

Jonas

Absolutely – I’m firmly convinced of that.
What made Germany strong in production over the past decades was a high degree of automation, driven by competitive pressure.
Now we’re seeing the same kind of transformation happening in knowledge work – in the digital, computer-based work we do every day.
At Murrelektronik, this same shift is happening on the shop floor – automation in production.
And I believe we’re just at the beginning of a comparable wave in the office world. That’s why we need highly specialized solutions, like Shopfloor platforms tailored to industrial needs.
Because that’s where the right use of technology truly makes the difference.
Foundation models like GPT are powerful enablers – but they’re ultimately interchangeable.
They’re becoming a commodity – they’re not the unique selling point.
What really matters is how you integrate that technology into your systems and processes, and what kind of real-world use cases you build on top of it.
And I’m convinced that in a high-wage country like Germany – where the pressure to be more efficient is intense – the combination of speed and result quality will make a huge difference in the years to come.

Fabian

From my perspective – even though, as the person responsible for IT, I don’t necessarily like to admit it – today almost anyone in our company can build and automate something using Excel.
I think in the next three to five years, it will be just as natural to create your own AI agent – tailored to your specific use case.
Excel will gradually be replaced, and honestly, I’m happy about that.
What worries me a bit, though, is the idea that suddenly everyone might have five different AIs running on their computer. Coordinating and orchestrating all of that will be a major challenge, but also a huge opportunity.
That’s why we’ll need a solid platform to integrate all these solutions properly.

Absolutely – that sounds like a whole new podcast episode in itself!
Especially topics like data protection and managing multiple AI tools definitely deserve more attention.
But for today, I just want to say: Thank you both so much!
I think today’s episode gave a really clear picture of your business case, your vision, the shifts happening in the market, and the kinds of challenges and demands your customers are facing.
Thanks for explaining it all so clearly and tangibly.
I’ll leave the final words to you – and once again, thanks for being part of the conversation today!

Fabian

Thank you very much for having us – I really enjoyed it!
Our project will definitely continue for some time.
We’d be happy to share an update once we’ve integrated more services.

Jonas

Likewise! Thank you for the invitation and the great conversation!
I’m looking forward to further exchanges and shared inspiration.

Thanks again, and have a great week!
Take care, bye!

Fabian
Thank you, bye!

Jonas
Bye!