AI Employees Are a Design Decision, Not a Hiring One

Welcome to part two. In part one, we uncovered in our interview with Oliver, founder of ADA AI, why the traditional back office is breaking under the weight of modern startup operations and how AI employees are already taking over full workflows at companies adopting agentic AI. But today we go even deeper in part two of this episode. Originally planners one episode, but Oliver was.

Oliver was giving so good and long answers, we spontaneously decided to break it into two. In this episode, Dr. Oliver Lugos reveals the technical breakthroughs that separates simple chatbots from real autonomous agents. The industries that are secretly becoming agent ready.

And why building an AI workforce might be the defining strategic advantage over the next decade. If you ever wondered how autonomous an AI should be, what God rays matter, or how far you can push automation without losing control, this is the episode you cannot miss. Let's jump into part two. Welcome to Startuprad IO, your podcast and YouTube blog covering the German startup scene with news, interviews and live events.

Welcome to part two of our conversation with Dr. Oliver Lugosch, co founder of ADA AI and previously co founder of the Razer Group based in Berlin, the E commerce company that scaled beyond 700 million euros in annual revenue. In part one, Oliver walked us through the origins of ADACOP, the rise of AI employees and why Agentic AI is fundamentally different from every wave of automation that came before it. Today we shift gears.

In the second segment, Oliver breaks down the technical and operational mechanics that allow AI employees to execute multi step workflows, make decisions, coordinate across tools and operate at levels that were considered impossible just a few years ago. We explored the real breakthroughs behind autonomous agents. How companies move from one agent to AI workforce, would guardrails prevent runaway autonomy, emerging industry leading the adoption curve, and how agentic AI will reshape high rank robos and organizational design.

In part one showed us why AI employees matter and part two shows us what's coming next and how fast. Oliver, welcome back. Thank you. Great to be here.

Totally my pleasure. For everybody listening, it was a few days or even longer for us, it has been just five minutes. So let us talk about what's the real technical breakthrough that makes AI employees fundamentally different from chatbots or those RPAs? Why we have chosen to call them AI employees is because that wording should show that the usage, the application, goes beyond these very singular purposes.

What do I mean by that? A chatbot is there to chat, Right? It's mostly confined to the exchange of communication. Right?

You can have small LLM helpers that for example, summarize documents. Right. And give you quick notes around it. But the thought about AI employees is that it can go much, much further, it can go beyond.

It can actually act on a longer process scale, so to speak, and connect many different dots. And what you need for that is many different things. You need the capability of using LLMs in exactly the right way, that they produce the outcome that you want human like communication. You need integrations into many different systems.

You need an integration into your communication system, into your Outlook or whatever email provider you use, into your maybe direct messaging service. You need integration integrations into your system of records, into your erp, your CRM, you know, for the logistics folks out there, into the tms, the wms, you need a human interface, you need a front end that the team can actually use, right? And there are all of these different elements that you need to combine in order to really think an entire process.

Because only if you think an entire process, you can also have impact that is really sizable, that is really substantial. And it's more than a little helper here, a little helper there, where yes, it might be interesting, it might be helpful in everyday life, but it doesn't create this measurable meaningful impact for organization that truly, you know, gives teams the room to breathe and feel like, wow, something has really changed. And because you can combine all of these different things now, and because you can tie all of these ties together, you can think about a broader automation and more impact that creates benefit for the organization.

As AI grows quickly, what patterns are you noticing in the demand of your customers? Which industries are or which processes are agent ready? Now, as of today, the pattern that. Emerges is that companies often have one or two use cases that is on top of everybody's minds because it seems to be present, it seems to be that people discuss these and we often start there.

And they most frequently lie in the procurement process, sometimes in the distribution process, but often in procurement. And then when you go deeper and you start to work on the project and you start to build, then more and more people hear about, you know, what is possible and what can be done nowadays with the help of LLMs and other technology. And then more ideas emerge and people start to see more use cases from different parts of the organization and start to discuss and also initiate further project, further use cases across the org.

So I would say classically your procure to pay and your order to cash flows are agent ready or automation ready? Really? I think the word AI agent is something that I find, I don't know, difficult in how to interpret it. Right.

What does an agent Actually do. What does an agent mean? I feel like it's a word that everybody uses. But what is really behind that, behind that curtain?

What, what, what is an agent? I would bet if you ask 10 people, you get 12 answers. That always reminds me, in the, in the mid-1990s, everybody was trying to sell multi multimedia computers and they had experts from like five new newspapers doing the elevator pitch and everybody talked about something different. Yeah, I fully, fully agree.

Right. And by the way, I'm not claiming that AI employee is the best wording, right. In any way. I'm not claiming that.

I'm just saying that, you know, these AI agents have, have popped up and I find it difficult to grasp what is really, really meant by it. In, in any case, I think these use cases pop up throughout the classical again, procure to pay, order to cash cycles also on the money flow, right. Speaking about finance, order invoice processing, issuing of invoices, reconciliation accounting, also these processes are ready. What you shouldn't forget is that this human in the loop process is something that you typically start from, right?

If you want to automate a process, do it, but still have a human confirm certain actions, confirm certain communication that goes out. So you can already get 80% of the efficiency of the time saving, maybe even 90%. But you still have that component of a human double checking and ultimately signing off on certain actions. And that is typically the starting point.

And then when you want to move to full automation, get these last 20 to 10% of, of, you know, capacity of efficiency. That's what you should do under certain circumstances, right? If you really have the trust, if you really have seen a lot of positive examples, a lot of perfect drafts and proposals from the AI, that's when you can move to full automation. But that's a very, very organic process.

That's a very trust driven process that just comes over time and again. Even if you are stuck with these, you know, human in the loop processes, you still have most of the value that will, that you will get ultimately. We've been always talking about the employees who get rid of very boring stuff. We've been talking about the entrepreneurs who get more efficient with a few employees.

But where are the customer wins? What's the most surprising outcome a customer has seen after deploying an AI employee? The first thing that customers think about is the efficiency, right? And saving capacity, saving resources.

But what also comes out of, of all of these projects is an improved speed of processing, data processing input, responding to customers, responding to suppliers, reacting to any kind of signal that comes in really. So you get rid of this backlog, right, that often piles up and it can again slow down other parts of the company. Other teams will notice, will notice that processes run quicker and more reliably with the use of these automations. And the quality of the output data is typically much better.

If anything, it's certainly of higher consistency, right? Because if you use an AI based automation, there's consistency of how things are interpreted, of how things are put into a certain place in the system, for example, and that consistency drives usability of that data because, you know, it's no longer, I would say a human spread of how things are being done, but it's typically very clean and very standardized. And that is a benefit that many companies don't foresee, but that they experience as they use the AI employees because they make the data more consistent and more usable going forward.

Also for other teams. I see you do that because you want to scale. But what are some scaling challenges with AI employees? What's the most underestimated challenge that companies, your customers face when scaling from one agent to an AI workforce, to many agents?

When we think about these automations and changing processes in companies, ultimately that's what we're doing, we're changing processes. Processes. Then you should never underestimate the human factor, right? You will always have those teams that are responsible for the processes, that oversee the processes, that further develop and advance the processes.

And you cannot go faster than the team can go, right? You always have to take everybody along and make sure that these changes happen at a pace that is digestible for the organization. And so the limitation in scaling is really not a technical limitation. It is not a limitation of, you know, being able to produce these automations.

It's a limitation of how quickly can teams and organizations adopt to these changes. And I think that's a very, very natural limitation. And it's something that you should always keep in mind when you think about how quickly do I want to scale from 1 to 2 to 5 to 10 use cases or AI employees. I see in the founders world we will be talking about what's the workflow.

No founder realizes can be automated, but every founder should automate using AI employee. Let's do a little bit outlook into the future. How do you see the balance between human teams and AI employees evolving over let's say the next five years? I know it's pretty, pretty early, but givens a positive vision, I think this.

View is pretty clear. I think there will be less and less and less of these painful Manual things that you have to do, the repetitive ones. I think we will spend more time on really figuring out stuff, figuring out how to do things and then having your AI employee, your assistant, your agent, whatever you want to call it, take over what you have defined. It will also help us become more creative.

You can use LLMs to get new ideas, get new impulses and then turn them into great ideas, turn them into great execution and really move your business, your startup, to the next level. So I think you will have less of this boring, painful piece and you will have more of the creativity. Do things differently and, you know, come into a state where it's, it's, it's exciting to advance your business because you no longer have to take care of the very, very basic stuff. I was just wondering about AI employees.

The very, very boring stuff. I was wondering, would it be a dream for you to automate or start automation? German tax authorities but because I do believe there will be quite a lot of potential and I also do believe there are a lot of people who have jobs that are very repetitive and I'm sure they would love to get rid of at least those steps of the work. I'm not a tax expert, but what I can say as, you know, being subject to tax, I fully agree with you.

I think there are probably many, many process steps in, you know, filing taxes, anything around taxes that could be at least simplified or accelerated with AI and LLMs. And I think there is a huge space that is to be tackled in that area. I fully agree now around the regulation aspect of it. I cannot speak to it, but I'm sure we'll figure this out and make sure that, you know, we can also use AI in these, in these topics and these areas of businesses agree.

Going back into the, from government into the real industries. We've been talking about AI agents helping employees to get rid of boring work. But I was wondering, how do you imagine agentic AI reshaping the org charts of larger companies? I think AI will start to become part of the org charts.

Right. Ultimately, that's where we're headed. And what does it resemble? Org charts basically resemble or mirror, you know, certain resources capacities.

I'm sorry, I'm not trying to be, you know, inhumane or anything. I'm just trying to abstract it. Right. What is mirrored in an org chart, it is, you know, human resources executing certain processes, certain tasks, taking over certain responsibilities.

As we see AI applications take over that role, they move into that capacity. Right. And you will see that a person will be completely occupied with running many, many, many operational or a large volume of operations through the help of AI, AI employees, agents, whatever you want to call them, and being that person that keeps an eye on all of these processes, right, keeps an eye on quality, further advances the process, further advances the model, the application. And really being able to have such a huge leverage, right?

One person, imagine one person handling, I don't know, thousands of customer requests every day. Why? Because they have LLMs and AI at their fingertips, right? And they can do this at a speed and a quality that was unprecedented, that you could never have had with just humans.

Because you can further improve these algorithms and these automations more and more and more. And I think that operational leverage that you can have will result in org charts looking very different from today. Do you know when you've been talking about the difference in org charts, what do you say triggered one idea in my mind because when I was working in a lot of different companies as a consultant, if you're a good consultant, you get invited to the Christmas party of the company and then you do have the retirees, the people who come in, they retired and they talk about what was work like 5, 10, 20 years ago.

And I was wondering if at one point you'll have like one room in a company. I had in mind a lot of old screens where you can basically talk to old AI agents and, and pick their brains on knowledge from the past or how it was done in the past. Why not? I think it would be enriching, right, if you could tap into that knowledge.

It's still fresh because it will always be fresh, right? It will remember and we'll be able to help us and learn, right. Obviously you will never be able to fully take over that human element, right? You will always have the core team, will be humans at your Christmas party and you will probably not celebrate with AI employees.

I don't know, maybe we do, I assume we will not. But having that, you know, this, these capacities of remembering stuff very clearly giving you insights, you can ask any questions it will hopefully answer in the correct way, in a true way. I think this is a value add that we are just starting to understand, just starting to grasp really what I. Had in mind when I was talking about this was having a digital room, not necessarily to be entered during Christmas party.

We all know that can get out of hands. A digital room where can basically look into different screens and have all the old AI agent archived there, which of course will be a big job for knowledge management. I'm sure you're working on IDs for that as well. Let's pick your brain a little bit about a contrarian view.

What is one unpopular truth about autonomous agents that nobody in the AI industry wants to admit? I think it comes down to this topic of autonomy that you mentioned before. I think the unpopular truth is that we are still far away from this true, true fundamental autonomy of really running human processes. People sometimes claim and continue to claim that we are there already.

I don't see it. I have not seen it in application and I think it's still a little bit of a way to go to really have that live and ready and in production. For our founders out there, for our audience. DM us on LinkedIn, me, your manager.

What would be the one process you'd automate first? We'll share the most creative ones anonymously. Looking forward to ideas there. Let's get a little bit towards the end and this pencil a little bit at end.

The advice, what's your advice to early stage founders who want to build AI ready operational structures when they're starting out like today or. Well, Today is late November 2025, but this will air in January 2026. A lot of activity is going on in January. Everybody wants to start something new.

What advice would you give to entrepreneurs now thinking about setting up the company? Great question. I think yes, it should be AI native, I think because not starting something AI native or with AI in mind would be foolish given all of the technology that we have and the capabilities the technology brings. But I won't, I wouldn't overdo it.

I wouldn't say, you know, first day you start thinking about AI. No, first day think what is your business? What do you want to do? What's the value that you will bring to your customers, to, you know, society, to the world.

And since post corona, how do you make money? What's your unit? Economics, your mix of scale. And then think about how I could raise that.

That would be something I will be thinking about. How about you, Oliver? Exactly 100%. So first start with, you know, what are you doing?

How do you create app value? How do you monetize? All of these are very, very important questions. And once you figure that out, or once you have an hypothesis and you want to test it, use AI as a tool, right?

Use LLMs to then scale yourself, scale your time. But if you're not a business, you know that, that, that sells AI where AI is really the core product. Don't start with AI the very first day. Start with the fundamentals of the business.

I think this is true. This will always be true. And once you've started to figure out things and you have your hypotheses or something that's even proven, use AI to scale. Use AI to create that leverage that you want to have.

I'm curious for one thing, because we go into the last standard question, and then we have the two mandatory. But before we go into that, can you see a future where the AI employees are targeted by hackers and used for nefarious purposes? I think that time is already here. I'm very sure that hackers are already trying to attack LLM systems, AI systems.

And it's something that you have to stay on top of and that you have to build countermeasures against. So I think that's absolutely critical. Whenever you design any kind of AI system that you have that in mind, people will always try to get to your system. They will always try to hack you.

So security is number one priority, particularly if you think in broad applications. Right. For a small, small company, small startup, they might not be the first target, but when you move to larger scales, then you're certainly targeted day in, day out. And that must be very core to how you approach the entire thing.

I see, I see. And the last question I prepared for this interview, if every company in the world would deploy AI employees tomorrow, what do you guess would be the one that breaks first? Like what function? What capability?

What would break first? Great question, great question. I mean, looking at what LLMs are great at, I would assume some of the numerical stuff would probably break. I don't know, maybe the forecasting piece or the optimization piece might be something that is most fragile because, you know, LLMs, large language models, are not designed for numerical analytical approaches.

But that's just my guess. I hope we will not see that happen. But if I had to guess that, that would be my take. Our standard questions are pretty, pretty normal.

Are you open to talk to new investors? We are always open to talk to new investors. We are not actively fundraising, but we are always open to talk. And this is a funny question for you, because are you also open to look for talented employees that you cannot yet, that you cannot completely replace with AI employees?

We are also looking for great people. Absolutely. At any point in time. So, yes, we are hiring.

So basically, we'll link down here your career website and your personal LinkedIn profile so investors can reach out to you. Perfect. Let's do that. Oliver, thank you very much.

Was such a pleasure to have you here twice. Thank you. Thank you, Johan. It was a real pleasure.

Was great talking to you. Same here. Have a good day. Bye bye.

Thanks. That's all folks. Find more news streams, events and interviews@www.startuprad.

IO. remember, sharing is car.