Daniel and Felix: Co-Founders at Cellbyte on $2.75m raise!

0:00Felix Steinbrenner:Hello and welcome back to the Scaling Europe show. I'm here with Felix and Daniel, two co-founders from Cellbyte. Cellbyte has just announced a$2.75 million round to cut pharmaceutical drug launch timelines with AI. Amazing news, guys. Congratulations. How are you feeling? Thank you so much. I'm feeling great. Daniel, how are you feeling?

0:20Daniel Moreira:Amazing. Amazing. I think that's like when we finally can focus now on really building the great platform. I think that's amazing.

0:29Felix Steinbrenner:Yeah, sounds great. Can you go into a bit more detail about what it is that you guys have actually built?

0:34Daniel Moreira:Yeah, happy to take this. So we built Cellbyte, which is basically an AI-native platform based on a genetic workforce with the mission, I think this is the most important thing, is to accelerate drug launches. And this is what we want to do with Cellbyte because after clinical trials are completed, it still takes a long time for patients to get the medicine. And this is what we're tackling with Cellbite to really accelerate those delays between one and two years to streamline this.

1:08Felix Steinbrenner:Amazing. And why this problem? Why this area? How did you get involved in the kind of, yeah, the world of pharma and drugs and all of this?

1:18Daniel Moreira:Interesting, interesting story. And I think Felix can add also the nuance on how this came about. I think many, many things came together to build Cellbite. I think when I started seven years ago after university, having studied business engineering and machine learning, I went into life science consulting. And shortly after being there, driven by curiosity, right? So to work on medicines being launched a couple of years ahead, I think was really amazing to shape the strategy. So I was there in consulting. But one thing stood out is that there was limited innovation after the clinical trials. We have seen a lot of innovation in the clinical trial space, but after that, really limited.

2:12Daniel Moreira:And one reason was that there is simply a lot of unstructured data. And pre-LLM, I think there was simply limited capacity to extract real insights out of this. So this work was heavily relying on subject matter experts, on anecdotal evidence, on massive teams researching, extracting, and analyzing lots and lots and lots of data. Because typically these regulatory bodies, they have to publish all these documentations, those review processes of drugs. So there's a lot of value in it, but really difficult to extract since they're lying on a semantic layer. So I think with the rise of large language model, this changed dramatically.

2:55Daniel Moreira:And while I was working in my project, I saw basically, I could not imagine years ahead, I'm still doing the same work as I was doing with the teams while I was using and trying out chat. I think at that time it was like 3.5. I basically had this epiphany. this will change everything. And this was the time then where, then I think Felix can jump on here, where I started the conversation. The link between us was my brother, who was at Felix, built Felix, the company with Felix before, which came from a completely different space. But then we sat together and said like, hey, we have to tackle that space because I think this is an overlooked space where there's lots and lots of potential.

3:40Daniel Moreira:And this is what we did at Sellwide.

3:43Felix Steinbrenner:Amazing. And look, you've raised$2.75 million. I think it's been led by Frontline Ventures, but also YC, Pace Ventures, South Capital, some angels. What unlocked this round? Was it the product that you've built? Was it traction that you've seen? Was it the team that you've built around you? What kind of gave the investors conviction that you guys were going to solve this problem? Yeah, I'd say what unlocked this round was actually our really strong traction. So we went basically from zero to 200K in annual returning revenue in a matter of weeks, which really proved that we identified a pain point in the pharmaceutical industry.

4:20Felix Steinbrenner:And that also large pharmaceutical companies were and still are willing to adopt AI in the pricier market access space when it comes to launching these drugs. I think also what Daniel briefly mentioned is our team composition. So he brings the pharmaceutical experience. experience. Our other co-founder, Samuel, has been working with large language models for many years already. So he's the best engineer that I personally know, I can tell you that. And yeah, I've previously founded another company, Last Mile Logistics, which actually was held locally and we pivoted locally to Salvite. So from Last Mile Delivery to pharmaceutical drug launches.

5:01Felix Steinbrenner:But yeah, they basically also show us that we have some grit, we don't give up easily, and we have a good team that can tackle this problem head on. Amazing. And that traction in the very early stages is phenomenal. Where did that come from? I can't imagine that pharmaceutical companies, they're often seen as very traditional, old school. How are you able to sell into those types of businesses? Yeah, so basically when Daniel told us about his problem, We conducted many interviews with drug or ex-professionals in the drug launch space to really understand what is it that they are currently struggling with.

5:37Felix Steinbrenner:And basically, we understood that there's a very fragmented and messy data landscape across regulatory information, health technology assessments, which is basically the step that comes after, which a pharmaceutical company has to do for every market individually to basically get market approval there, and then the pricing reimbursement part. And what we have basically done is used AI to scrape a lot of data, clean it, and interconnect all of these together to help pharmaceutical pricing and market access managers grow our insight from this messy, fragmented data in just a couple of seconds. Phenomenal.

6:15Felix Steinbrenner:Very cool. And, you know, how's your journey been, I guess, from sort of going through that pivot? But yeah, how have you found the process of kind of adjusting from the last mile world to the farmer world? Yeah, it's been quite a steep learning curve for me, for sure. I think it's quite a different market, but we were able to draw from many learnings back then from last mile logistics. We use AI a lot for route optimization and for forecasting demand, basically. And it helped us a lot in generating insights from this large amount of data in logistics back then. We can now do this in the farmer space.

6:51Felix Steinbrenner:And our mission in the end is to get drugs to patients faster. So I think we really have this chance here at the pharmaceutical industry to accelerate drug launches, which in the end helps patients get access to treatments that they previously would have gotten not at all or maybe delayed. So this is really motivating for me personally to have this real-world impact and also this huge vision to work towards in the long run. I mean, it's an amazing mission. does that mission help? You know, the first startup I worked at had a good social cause at heart. Does that help you when it comes to moving quicker, hiring talent?

7:31Felix Steinbrenner:Is it something that you lean quite heavily on? Because a lot of people do want to do work that has a positive impact. Yes, definitely. I think it just really helps us get top talent attracted or basically interested in what we're doing. So I think this can give us an edge also when it comes to the goal that we're all working towards. I mean, Daniel, you can maybe add your thoughts in this one.

7:55Daniel Moreira:I think one of our engineers who joined us was so impacted by this because he has a girlfriend who has a rare disease, right? And he had, basically in his close environment, he saw how the impact is if you have to wait for certain treatments to be ready to be used. And I think this is really something that can be changed, right? That needs to be changed. Just to give you an example, like a very, very simple one. The FDA recently published rejection letters, like why certain applications were rejected and not able to launch. And some of the reasons, right, were simply due to, I don't know, mislabeling, wrong information that is really easy to have.

8:48Daniel Moreira:Like, imagine, like, you're preparing to launch a drug, like, billions of dollars. And somehow it fails and you have to go into, like, a resubmission due to really, really, yeah, aspects that have nothing to do with the billions and the clinical trials that have spent, right? So simply these terms, if you can run your submission documents across what have been rejection reasons already in the past, this streamlines already all your documentation processes. So on a scale that we're building globally with many, many national buddies, this is what we're doing in Sellby to really, really streamline that work.

9:33Felix Steinbrenner:Yeah, I mean, it's ridiculous that you're telling that story. Yeah, it's to waste all that time and money on, I guess, what is just like administrative errors or something. I want to talk about the team and how talent itself more broadly. You know, you're building in Germany, you're building an AI company for the pharma industry. How are you thinking about team structure and what top talent looks like? Are you trying to get people with a lot of contextual experience? So people in the pharma world, a bit like yourself, or are you trying to focus more on AI experts, amazing engineers from some of the best labs or frontier models?

10:09Felix Steinbrenner:How are you thinking about top talent and specifically within Germany? Yeah, so maybe first of all, we're actually hiring all the talents directly here in Munich or at least expecting everyone to relocate here because we believe it's something else if we are all in the same room basically working on this hard problem together. We are right now still a small tip, So we have made six full-time hires so far, or that is actually including us three as co-founders. And at the moment, we're focusing on finding the most outstanding talent to join us with a focus on data acquisition engineers that can help us really broaden our coverage to reach this global scale completely.

10:49Felix Steinbrenner:then also AI engineers that can help us build these agentic workflows specifically for the pharmaceutical industry and basically to understand how these workflows have to be built. We are also hiring go-to-market people that would be pharmaceutical experts in certain workflows that have worked in the industry for a while.

11:10Daniel Moreira:But it's challenging.

11:13Felix Steinbrenner:Yeah, I love talking to companies and founders like yourselves who are building very it's like niche verticalized AI because it's such a hard combination to get right of really top quality AI talent but also capturing that contextual expertise that enables those engineers to then build the products that serve that niche of that industry and I think it's really interesting because if you get that right and if you do kind of marry the two together you've got amazing defensibility because it's so hard to replicate those two amazing pools of talent um you know looking forward you know you raise some money now i imagine you're like heads down building shipping selling what do you want to achieve what is it that you want to do to kind of get you to that next stage of the business and maybe it's the next level of fundraising maybe it's market expansion you know what is it that you need to achieve to be able to be like okay we're done with this stage on to the next one

12:06Daniel Moreira:I think how we're building Cellbyte is, in particular in pharma, you have to do things 99 % 0.9 % correct, right? There is no 85 % solution that does the job. So everything in pharma specifically needs to be top-notch and fully correct. And you have like one shot to deliver. And with this in mind, we have the last couple of years focused really on building a very strong data foundation that really connects all the relevant data sources in the pharma space to build robust applications on top. So this really enables us to test applications and run them robustly with pharmaceutical teams. And I think with this, having a really strong data foundation, strong applications running on top, you can connect those to end-to-end workflows.

13:07Daniel Moreira:And this is how we're building Cellbytes. So we're moving from the data foundation. You're working every day on the data foundation. We're working now every day on the application layer. And together with the teams and our user base, we're exploring what are the key workflows that they want to attach. And our goal here is really to build the best in class applications that really solves business problems. Because I think we're approaching all these things not just like you have data providers, you have consulting companies, you have the pharmaceutical players in there, you have the teams in there.

13:42Daniel Moreira:So many different worlds right now, but we're basically taking a step back, zooming out and thinking, OK, what is really the issue that needs to be solved where AI can really deliver robust solutions? And this is what we're tackling together with the users. So we have we run a lot of interviews. We have a lot of discussions to really build strong solution. I think this is the goal and this is the yeah, our our path to growth is really building on this. we can really build the key tools to solve those business questions and grow with that.

14:15Felix Steinbrenner:Amazing. And what are the metrics that you look at to know that your product is working for your customers? So what are the metrics that you're thinking, OK, we have to get this quicker, we have to get this more accurate? What are those metrics and how are you working towards improving them? I mean, we're monitoring basically user behavior and speaking also a lot to users. So it's both quantitative data that we're evaluating. We're also benchmarking constantly different AI models with regards to how we can use them for data extraction and analysis. But then also, for example, in-science extraction that users can do directly across tens of thousands of documents at the same time.

14:59Felix Steinbrenner:And then we're basically on a daily basis speaking to our users, trying to understand how they're currently using the platform. Quite often we're also shadowing them how they use it because I think it's very important that you look your customers over the shoulder. And if something doesn't work, then kind of have to sit back quietly and cringe at how they're using the platform. But then it really helps us understand, okay, maybe we need to change something at this point in our product. but yeah from a financial perspective it's month-over-month growth that we're looking closely at of course and revenue targets yeah i think you know you mentioned that it kind of in this space it's quite similar to the compliance space in the sense that you can't just ship like a an mvp you sort of you've got you've got to build a product that is robust enough to do what it says because you know you can't have holes or gaps or small mistakes because if you do it it doesn't work how did you find like what was your approach i guess in building the product to that point were you trying to get feedback throughout that process of getting it to kind of the minimum it could be or did you have to build something entirely by yourselves before

16:06Daniel Moreira:you could even show customers what did that look like yeah i i think we built um we talked a lot first um and also after they like we have we we can draw back on um years of experience in that space, right? I think this is key to first, like, being somehow respected in that area, because you speak the same language, they know you can solve these problems in a manual way. And now we're translating this into a product. So deep domain expertise, I think, is a prerequisite here. And then you have the additional capability you need is to translate the challenge that you you have to solve manually into a product, right?

16:48Daniel Moreira:Because this is where the gaps increase, where you basically can get with AI to 80 % and then the rest to 20 % that a human needs to close is actually the thing that you have to nail and it's super, super difficult. So we really like through all the interviews, discussions, we were basically trying to find like, what is the end-to-end process that we can build a product on where you can use them and it's super robust, right? So I think this is how we started. And the beauty about the pharmaceutical space is that pharmaceutical companies, they value. The value of the product matters more than the price, right?

17:34Daniel Moreira:So you can really focus on building something and this is something that will generate some value on the pharmaceutical side and we will respect that you will generate value if you generate value for them you can also generate value for the company and I think this is definitely something that we have to go and before we launched we had smaller pilots teams that were looked into users that were really open and really excited about that space. So these are our, like, I think we're still working with one of our, like, first heroes that they just love that space and they see how much you can do. And this is, you have to really find those people that believe in you and believe on the solution, even if you're not seeing already what is going to happen, how it's going to change.

18:27Daniel Moreira:So once we got those, I think this was really key to win then the full team. And now, like, after winning the teams, We're expanding to other teams, right? Because this is, I found out also a very strong word by mouth spreading.

18:43Felix Steinbrenner:Maybe just one small addition to this. We actually focus on certain markets in the beginning. So we started with the German market and IMA from a regulatory perspective. So then basically draw on the learnings and build more markets by the side of this. Got it. Okay. So look, two final questions. I think I'll ask one each of you, but it's all about the highs and the lows of the journey so far. uh so phelix why don't i start with you what has been the hardest or lowest moment you know the time where you've had to like really push down maybe something has gone wrong or broken where you felt like okay maybe we're doing the wrong thing um or there was a really hard decision that you had to make something where you were really at the bottom of your mood and you had to kind of push through i would say there was probably around christmas one year ago where um we actually had to make sure that the German data that we had back then, which was more or less our only database on the health technology assessment perspective, had to be super robust for our customers.

19:43Felix Steinbrenner:So I was actually going manually through thousands of lines with, I don't know, probably like 25 columns, needing to make sure that every single cell was completely correct and up to date. And there was an inherent logic in this. And I was at this moment thinking, okay, we need to make sure that in the future we can utilize AI to help us with this analysis because if I have to do this for every database manually this won't work um so luckily we figured out how we could use AI and still have a human in the loop um but yeah that was probably uh quite a quite a low for me amazing what about um what about the high point you know this view Daniel like the highest point of your journey where you thought okay this is going to work this is going to happen something clicks maybe a funding around maybe a customer maybe you know you'll you see your ai working in the real world what was it

20:34Daniel Moreira:where you're like okay let's go this is it um i think one particular example that it stuck like stays still in my mind is um when i had an a call with with someone in the us and i think we were just like five minutes into the call i was just like opening the platform like i just asked him a question and he asked me like, hey, can we, I'm actually really on this problem right now. Can you look up for me if it was like about the clinical trial design and he wanted to check how Germany could perceive this? Five minutes. We just put in a few sentences, checked it, and he basically said like, that's a deal.

21:21Daniel Moreira:This was like so game-changing for him to have access to these German insights in a robust way. And I think there we know, like, okay, we are really into something that can generate value across the globe.

21:38Felix Steinbrenner:That's amazing. That must have been such a phenomenal feeling where you're like, wow, this is exactly what I've been dreaming of. Well, look, both of you, huge congratulations. The list of what you're building is amazing. And the fundraising is only going to take you further. So thank you so much for joining me and best of luck. Thank you so much, Seth. See you next time.

21:57Daniel Moreira:Thank you so much. Thank you. See you soon.