What does a heavy metal album cover have to do with biotechnology blueprints? According to Joaquín Peña Gazal, Stämm’s Graphic Designer, the answer lies in the radical versatility required to solve visual problems. In this conversation, Joaquín demystifies the often-overlooked role of a designer within a deep-tech company, describing his work as that of a “visual translator” who bridges the gap between complex raw data and public understanding.

From the real-world dangers of bad communication — illustrated by a misleading billboard that could literally drive you off the road — to the philosophical divide between Art and Design, this interview offers a look at how creativity powers scientific innovation. Read on to discover how graphic design is shaping the future of how we see and understand biotechnology.

Joaquín: There’s something I get asked a lot when I say I’m working at a biotech company: “Oh, but why do they need a graphic designer there? What does a graphic designer even do?” And well, you already know I do a ton of work here. So I highlight the interdisciplinary nature of design. Because people say, “Oh, graphic designer,” and think books, posters, stickers… but you could see one collaborating on an architectural project just as easily as in a biotech company. There’s always a need to visually communicate something. Whether it’s something very specific, like movie posters — going back to that — or cutting-edge tech blueprints like we work on here. Blueprints, or showing certain aspects of our technology graphically to make them accessible to everyone, or translating it into a common language. Sometimes I feel like my job is being a visual translator of sorts.

Do you feel you manage to communicate to the outside world what a graphic designer does in biotech? What do you emphasize when you answer that question?

When I answer that, I sometimes start detailing my tasks at Stämm. From helping with tech infographics to doing illustrations and designs for social media. But I think what I emphasize most is facilitating complex information for a lay audience. So, I either end up talking a lot about our social channels, or I talk quite a bit about the infographic work. Like helping break down information into different graphics where we pull out various items to show in presentations. I mainly emphasize those things.

How did you get here? Because you are a graphic designer in biotechnology as we speak. What did your training involve to be able to do this now? What did you study? What was that process like?

Well, something they tell you starting in university is that you’re never going to be “wedded,” so to speak, to one theme or concept. The graphic designer is kind of… maybe this isn’t the word, like a mercenary, right? (It’s the actual etymology of freelancer). You take the job that caught your eye a little, even if you’re not a super fan… A specific example my professor used to say: “You might not be a fan of metal, but if Iron Maiden comes asking for an album cover, are you going to say no to Iron Maiden?” So, you’re going to dive into that theme and learn a lot. A designer has to be someone very curious so they can dive into different topics, study them, and start designing with that. A design base can’t exist if you don’t get into what you’re seeing. There can never be just an external look and that’s it; you have to internalize all these things. It’s always about versatility, right? It happened a lot in college: they threw different themes at you, didn’t matter what, and you had to design. You had to make books, pamphlets, brochures, or infographics… you had to do it all, regardless of the topic. And then I got here… I always had an interest in biology. In fact, when I was a kid, I wanted to be a marine biologist… biology just grabbed me. When I started here, it was nothing like I imagined. It’s a level of biology that’s quite advanced, cutting-edge, innovative — at first, it looks like reading ancient Greek. But then, as you relate to more people here and start seeing what they do, you begin to understand. Besides the versatility and curiosity a designer needs, interdisciplinarity stands out. Without it, you wouldn’t be able to understand a lot of things.

Sure. Like, one thing is being curious, another is being a biologist. In the sense that you can be interested in the discipline, but you’ll necessarily have to talk to a specialist or a…

Exactly. Yeah, that happens to me a lot, for instance… I’ve been here for a little over two years. Sometimes they talk in biological technicalities that I don’t understand at all. But then, hearing them speak, I rescue different concepts, I pick up on different methods they use to refer to certain elements, and I design with that. I might not necessarily know the detail of everything I’m designing, but I do understand a general concept thanks to listening and paying attention to what’s happening in context, so to speak. You definitely have to lean into that context, otherwise, you end up doing something that isn’t design. Design is something thought-out, it’s methodical. Even if it seems purely visual and practical, it also has a lot of theory. So if you don’t have a theoretical base or a defined concept of what you’re doing, there is no design. So yeah, even if one doesn’t know specific terminology, there are general biology concepts that one picks up just by chatting with people here.

And what do you consider to be most important? Because okay, themes can vary, formats can vary. What seems most important to the eye or the training of a designer? I mean, what does the designer have to bring to the table? Because you can change the format: today it’s a poster, tomorrow an album cover, the day after a PPT. And the theme too: today biology, tomorrow physics, the day after progressive rock. What is the constant that drives design then?

The constant for a designer across all these different themes is facilitating information. Because the designer exists to solve problems, essentially. If you are involved in all these themes the common denominator is that they all have a problem communicating it visually. So the designer seeks to translate that niche information from that theme into something much more general. So it’s always about translating information. It’s always about showing a particular point of view, finding an angle to visualize something that might otherwise seem entirely theoretical.

With your experience as a designer, what is the hardest part of the profession or the hardest part of your role here at Stämm? The biggest challenges you’ve found?

I think the biggest challenges here at Stämm were finding a way to make all these professions I’m working alongside converge. Because you have other designers, industrial ones for example; you have sales people for an innovative product, marketing; you have scientists. So, all those perspectives have to converge in one place. And the designer is putting those points of view, placing all those viewpoints graphically, side by side, visibly, even merge them. Whether through a drawing, a schematic, or a photomontage; you’re showing all these perspectives and you have to make them coexist in the same design. Maybe there’s something that seems, for example, aesthetically cool and one side really digs it — like, the industrial design team likes how certain arrows or compositional layouts represent things — but then marketing says: “No, well, but it would be good to show this”, as they point on the opposite direction. And then, at the same time, the scientists make some clarifications about that information, and you have to figure out how to show that data while also being on par with marketing and maintaining the aesthetic that the other designers liked. That negotiation of perspectives is, for me, the hardest part. Uniting many gazes.

On the flip side, on the positive side, that’s the most pleasurable and fun thing, to put it simply, about being in this environment as a designer?

The most fun thing about being here is that there are many things the rest of the world hasn’t seen yet. They are quite new, whether in the shapes of the elements or products we make, I find a certain gratification in seeing it first and also editing it first. When they ask for something more complex like a product description via an infographic, it feels like I’m contributing to something a bit bigger than myself, because I know these things can endure and make an impact. And being among the first people iterating on it is great. I usually think: “nobody has ever seen this design before.” I don’t want to sound arrogant, right? I’m not saying I’m making the first lightbulb here. But it feels like that, like you’re working on something very, very new. It’s partly uncharted territory, and it’s exciting to get into that. That is partly one of the reasons I joined Stämm; I thought the premise and the innovation it brought were very interesting.

What did you encounter here at Stamm? I feel like science sometimes lacks a bit of graphic design, of visual communication. Do you have a theory on why it lacks that now that you’ve worked here? Or why that difficulty exists? Why it’s so hard to transmit scientific knowledge in this way?

Yeah. Something that happened to me is that when I first got here, some things looked very “textbook biology.” That kind of aesthetic. A lot of text, a couple of illustrative images, and that’s it. It was just that. I see that in a lot of other biology stuff too. It’s like they always stick to the literal. In biology, everything is pure data, raw data, which has to be shown literally. So sometimes they err a bit on the side of, well, being very obvious or not looking for lateral visual solutions — which is what a designer does, always playing with lateral thinking. So, if I’m going to show you a concept, I’ll try to make sure that concept is never too obvious or literal. But the thing with biology is, if it shows you a cell, it’s going to show you that specific photo of the cell. Right. Now, you can also do different things around showing it obviously, so to speak. You can alter the aesthetic a bit more to make it more attractive, more pleasing to the common eye, interesting.

Does graphic design imply any risk? Do you think bad design can be dangerous in some sense, or bad in some sense?

In my training in a couple of subjects, and also in real life, I experienced that firsthand. I think the freshest example I have is once we were driving on the highway and there was a soda billboard that was a project left half-finished. You could see they were building it and stopped halfway. It told you the mileage to “more freshness” or something like that, implying you were reaching a stop where there was soda. But of course, that sign was pointing — via the bottle, via the shape of the soda— in a direction where if you kept going straight that way, the road ended and you were going to skid off and get hurt. So we got lucky that it wasn’t finished, and we kept driving straight and said, “Whoa, crazy that it was pointing in that direction.” So that made me reflect a bit on how if we get the way we communicate wrong… we are generating quite a risk. Whether directly, like causing a crash due to a lack of logic in signage, or unintentionally conditioning incorrect things. Design is rooted in communication. So the dangers of design are the dangers of communication. You are always playing with people’s attention, in a way. Depending on how you reach those people, you can condition them incorrectly and cause harm.

Do you have concrete designers as a reference? Some pictorial references that serve as a guide for you?

I was always a big fan of grids. From the first time I saw them, I fell in love with compositional and modular grids. So I love Müller-Brockmann. He’s one of the fathers of graphic design, actually. I really like him. I also like how typography coexists with the image, so I like (David) Carson quite a bit too. Those are my two big references. There’s always going to be something with typography because I’m a huge fan of editorial design. I think that letters, their morphology, is quite a bit more expressive than they sometimes appear to be. Reading a word isn’t just understanding the concept the word says, but there is quite a bit more in the background in the way it is written; different meanings can jump out for the same image. And well, since I mentioned meanings, one of the reasons I’m a designer is also semiology. I am an extremely big fan of Barthes. Saussure’s linguistic sign, the interpretation brought to the image by Barthes, all of that is one of the reasons I said: “Wow, how much theory and how much meaning, how many concepts one can pull from just one image, from just producing an image.”

And how does graphic design get along with ambiguity? Perhaps regarding context, because design is always situated, right?

Well, that’s interesting because therein lies one of the very key differences between art and design. In art, you aren’t necessarily painting for a target audience. You don’t have a defined target. Your user is yourself. So you are expressing yourself. And while you might have a radical, “unique”, concept of your work, it’s going to be subject to people’s ambiguity because you aren’t making art for those people, you’re making art for you. Now, if you are thinking of a person, a specific type of person, a specific client, and the reproduction of that image for that type of client, it stops being art, it’s design. Because you are being methodical. You are planning. That’s the Andy Warhol debate, right? He was a designer but also an artist. There was quite a debate on that side. To what extent his work was considered design if, at the end of the day, there was a lot of art.

What are those characteristics that detach art from design? What mind, what considerations does the designer have that the artist doesn’t, and vice versa?

The simplest one to explain is that the artist doesn’t need to put themselves in the shoes of a person asking for a job. Both are open to freelance, right? And both can be commissioned, but the artist’s imprint won’t necessarily be altered by who asks for it. Whereas the designer, yes. The designer is going to study you. They’re going to learn who you are and how you see the world to be able to reproduce what you have in your head. The concept you want the designer to reproduce. So one is going to study their client, and the other one not necessarily. The designer sometimes is a medium, so to speak. Because you are translating the client’s ideas, but you look for it to be in their imprint, the client’s, not necessarily yours. For example, I came to Stämm and I have aesthetic preferences, but when you enter at a company, when you arrive in a client’s territory, you are going to limit yourself to their graphic rules. For example, at Stämm, we have a brand book. So I get here and the first thing I did was study the brand book to be able to speak in the same language as my client. And be able to represent him with the ideas he wants to show. The artist isn’t necessarily going to do that; the artist gives their own interpretation of the world.

Generally I ask everyone about the future in this series of interviews. Imagining Stamm’s future, everything we thought of. How do you think graphic design fits into a more mature, complete, and extended Stämm? What things are left to explore from graphic design at Stämm? What things would you like to see as a designer from Stämm?

It’s a very interesting question. Because sometimes it’s hard for me to be in a company with quite a lot of innovation on top of the product and such, sometimes it’s hard to project it. But accompanying it into the future, I imagine taking it all to a more mass-market level. I still feel like we are sometimes confining ourselves to the niche, except maybe in social media. But taking it to the masses, walking down the street and seeing an ad for our bioprocessor. Or seeing a TV commercial. Or doing a huge photographic production, not of the development itself, but of the finished piece with actors and everything for that commercial. Things like that. To keep exploring with renders. Something very attractive about this type of company is concepts turned into images. So working with renders is always like looking quite into the future, it’s like peeking through the keyhole of Stämm’s future, the renders. I think what I’m most enthusiastic about is seeing that render turned into concrete material. And seeing how I, through either photomontages, image editing, posters, presentations, and flyers seeing myself representing it and facilitating it to a mass audience.

There’s something about seeing Stamm’s imprint in other media that we haven’t explored until now. Like advertising, posters on the street, takin it into a next level. Taking it to, I believe, the common eye. Because we are still under the niche of the scientific bubble, but it is like very deep inside the scientific area sometimes. So having a person who isn’t necessarily involved in biology seeing the product realized and edited with a designer’s filter, it would be quite interesting to see how they react to something like that.

Is there something you would like to mention that we didn’t mentioned or haven’t talked about?

I think I spoke quite a bit about some of my interests like typography. I do have one more origin story as a designer that I didn’t speak about. I was always a big fan of comics, Japanese comics, manga. And there is an author who always inspired me a lot, not just for the story, but for the way he saw the medium, who was called Kentaro Miura, the guy from the manga Berserk. He composed the pages and made the drawing coexist with the grid in a spectacular way. You could really feel the dynamism, the movements of the drawings without them actually moving. Even if drawing is static, but because of how he used grids, how he used composition, he gave it a lot of movement. That, with semiology’s theory, were two things where I said: “Wow, I want to produce things like that, I want to make things that generate impact.”

Curious to see the result of this “visual translation”? Dive deeper into the technology that inspires these designs by exploring our detailed product portfolio at stamm.bio. And if you are hungry for more behind-the-scenes insights from the minds driving the biotech revolution, be sure to check out the rest of our interviews on our Medium page.

Central to this evolution is the widespread adoption of Single-Use Technologies (SUTs), often referred to as disposables. These systems, typically composed of sterilized plastic components and intended for single use, have become integral across the entire biomanufacturing process, from upstream processing (USP) through fill and finish.

Economic and Operational Drivers

The shift toward SUTs is heavily supported by compelling economic arguments, particularly regarding reduced infrastructure costs.

SUTs substantially decrease initial Capital Expenditures (CapEx) for facilities and equipment. In comparative studies, investment costs for single-use facilities were found to be approximately 27% lower than for their multiple-use counterparts (Eibl & Eibl, 2019). This advantage is primarily gained by eliminating the need for complex hard-piped infrastructure associated with cleaning-in-place (CIP) and sterilization-in-place (SIP) systems. SUTs shift these fixed costs toward operational expenditures, offering manufacturers financial flexibility and mitigating risks associated with early large-scale investment.

Operationally, SUTs provide enhanced speed and agility, which is crucial in the competitive market. They facilitate rapid changeover between batches and product campaigns, maximizing facility utilization, especially in multiproduct environments. This capability is particularly vital for contract manufacturing organizations (CMOs). Furthermore, using sterile, closed disposable assemblies significantly mitigates the risk of microbial and cross-contamination between product runs, thereby enhancing manufacturing quality assurance (Barnoon & Bader, 2008).

Sustainability and Efficiency Gains

Although SUTs generate solid plastic waste, their overall environmental impact can be lower than traditional stainless steel (SS) facilities due to reduced utility consumption (Pörtner, 2024). For example, eliminating CIP/SIP processes drastically lowers the demand for high-purity water, such as Water for Injection (WFI), which can represent a significant portion of a facility’s wastewater streams (Eibl & Eibl, 2019).

Comprehensive life cycle assessments suggest that end-to-end continuous processing using SUTs can reduce plastic waste by up to 57% and carbon dioxide (CO2) emissions by up to 54% compared to intensified fed-batch (IO-FB) processes in multiproduct facilities (Partopour & Pollard, 2025). Additionally, innovative approaches, such as on-site sterile 3D printing of single-use plastics using polylactic acid (PLA), demonstrate the potential for a substantial reduction in CO2 emissions compared to conventional petrochemical-derived single-use plastics (Achleitner et al, 2024).

These benefits are amplified when SUTs enable modern process advancements. SUTs are foundational for implementing process intensification and continuous manufacturing techniques, which further enhance efficiency and consistency.

Applications Across the Manufacturing Workflow

Upstream Processing (USP), the initial stage of bioprocessing that involves preparing and cultivating cells or microbes in the bioreactor, utilizes SUTs for nearly every unit operation. This includes single-use bags and container systems for sterile storage and preparation of large volumes of media and buffers.

At the heart of USP are Single-Use Bioreactors (SUBs), which come in many different mechanically driven types (stirred, wave-mixed, orbitally shaken) or are pneumatically driven. Stirred SUBs systems are becoming the industry standard, replacing traditional stainless steel tanks for new facilities. Wave-mixed bioreactors were an early milestone in disposable adoption and are widely utilized for seed train preparation and process development, offering gentle agitation characteristics particularly suited for shear-sensitive cells. Miniature systems are crucial for high-throughput screening and early process development, providing a reliable scale-down model. Continuous systems without agitation, as Stämm’s high-throughput bioreactor, represent SU systems that don’t even need stirring.

Downstream Processing (DSP), the series of purification steps used to separate and refine a desired product, has historically lagged in SUT adoption, primarily due to the technical complexity and high cost associated with purification.

However, disposable components are increasingly common throughout DSP. This includes disposable depth filters for clarification and initial cell removal and fully disposable flow paths for Tangential Flow Filtration (TFF), Ultrafiltration (UF), and Diafiltration (DF), which are critical for buffer exchange and concentration. Membrane adsorbers and pre-packed columns represent significant SUT implementations in chromatography, particularly for polishing steps and virus removal, offering high throughput independent of flow rate limitations typical of resin beads.

In the Formulation and Fill & Finish stages, SUTs enable closed systems for handling, storing, and freezing bulk drug substances using specialized bags and integrated manifolds. Single-use filling flow paths are critical for aseptic operations, simplifying preparation and cleaning validation for automated filling equipment.

Challenges and Future Direction

Despite significant advancements, SUT adoption faces several technical and commercial limitations. The migration of chemical compounds from plastic materials, termed Extractables and Leachables (E&L), remains the primary risk and regulatory concern, potentially affecting product quality, stability, and patient safety. Rigorous testing and validation protocols are required to mitigate this exposure.

Furthermore, while overall CapEx is reduced, the cost of consumables — especially for complex, high-volume bioreactor bags or large mixing vessels — can offset the initial savings, making lifecycle costs for complex single-use systems higher. Scalability is also constrained; practical size limits often restrict single-use bioreactors. Finally, the lack of standardization among various suppliers creates dependency issues and complicates the interchangeable use of components.

Future innovation is geared toward solving these challenges and capitalizing on the foundational role of SUTs in advanced manufacturing. This includes developing improved disposable sensors for comprehensive real-time Process Analytical Technology (PAT), creating better disposable downstream purification systems, and implementing flexible, modular facility designs enabled by fully closed single-use trains.

Ultimately, SUTs are expected to continue driving modernization, enhancing the efficiency, quality, and accessibility of biopharmaceuticals worldwide.

References

Achleitner, L., Frank, A.-C., Mesef, O., & Satzer, P. (2025). The future of single-use plastics in life science: Sterile printing of PLA reduces greenhouse gas emissions by 80% and enables carbon neutrality. Green Technologies and Sustainability, 3, 100132.

Barnoon, B. I., & Bader, B. (2008). Lifecycle cost analysis for single-use systems. BioPharm International, 20(7).

Eibl, D., Peuker, T., & Eibl, R. (2011). Single-use equipment in biopharmaceutical manufacture: A brief introduction. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 3–11). John Wiley & Sons.

Eibl, R., & Eibl, D. (2011). Single-use bioreactors — An overview. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 33–52). John Wiley & Sons.

Eibl, R., Löffelholz, C., & Eibl, D. (2011). Single-use bioreactors — An overview. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 33–52). John Wiley & Sons.

Farid, S. S. (2011). Evaluating and visualizing the cost-effectiveness and robustness of biopharmaceutical manufacturing strategies. In G. Subramanian (Ed.), Biopharmaceutical Production Technology (Vol. 2, pp. 717–741). Wiley-VCH Verlag GmbH & Co. KGaA.

Galliher, P. M. (2018). Single use technology and equipment. In Biopharmaceutical Processing (Vol. 5, pp. 557–578). Xcellerex, Inc., A GE Healthcare Life Sciences Company.

Galliher, P. M., Hodge, G., Guertin, P., Chew, L., & Deloggio, T. (2011). Single-use bioreactor platform for microbial fermentation. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 241–250). John Wiley & Sons.

Jossen, V., Eibl, R., Broccard, G., & Eibl, D. (2023). Single-use systems in biopharmaceutical manufacture: State of the art and recent trends. In R. Pörtner (Ed.), Biopharmaceutical Manufacturing (Vol. 11, pp. 3–40). Springer Nature Switzerland AG.

Kubischik, J., & Schaupp, M. (2011). Single-use technology for formulation and filling application. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 105–112). John Wiley & Sons.

Liderfelt, J., Rodrigo, G., & Forss, A. (2011). The manufacture of mAbs — A comparison of performance and process time between traditional and ready-to-use disposable systems. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 333–340). John Wiley & Sons.

Luitjens, A., Lewis, J., & Pralong, A. (2011). Single-use biotechnologies and modular manufacturing environments invite paradigm shifts in bioprocess development and biopharmaceutical manufacturing. In G. Subramanian (Ed.), Biopharmaceutical Production Technology (Vol. 2, pp. 817–858). Wiley-VCH Verlag GmbH & Co. KGaA.

Peuker, T., & Eibl, D. (2011). Biopharmaceutical manufacturing facilities integrating single-use systems. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 145–158). John Wiley & Sons.

Pörtner, R. (Ed.). (2023). Biopharmaceutical Manufacturing: Progress, Trends and Challenges (Vol. 11). Springer Nature Switzerland AG.

Rader, R. A., & Langer, E. S. (2019). Single-use technologies in biopharmaceutical manufacturing: A 10-year review of trends and the future. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (2nd ed., pp. 193–200). John Wiley & Sons.

Ray, S., & Dolan, S. (2011). Single-use virus clearance technologies in biopharmaceutical manufacturing: Case studies. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 311–322). John Wiley & Sons.

Riesen, N., & Eibl, R. (2011). Single-use bag systems for storage, transportation, freezing, and thawing. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 13–20). John Wiley & Sons.

Seeger, A., & Estapé, D. (2011). Production costs in biotech facilities: Single-use versus multiple-use equipment for antibody manufacture. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 351–361). John Wiley & Sons.

Tappe, A., & Gottschalk, U. (2011). Single-use downstream equipment. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 91–104). John Wiley & Sons.

Werner, S., Kraume, M., & Eibl, D. (2011). Bag mixing systems for single-use. In R. Eibl & D. Eibl (Eds.), Single-Use Technology in Biopharmaceutical Manufacture (pp. 21–32). John Wiley & Sons.

For Iñaki Laborde, responsible for the environmental area, the answer lies in thinking in a cyclical way. In understanding that every product, service or process has a reason for being… but also an impact.

In his role in the Quality Operations team, Iñaki combines sustainability and standardization to rethink how biotechnological systems are designed and operated.

We talked to him about life cycle analysis, eco-design, circular economy and how these tools can have a real impact on the development of cleaner and more responsible technologies.

Q: What is life cycle analysis and why is it important for the biotech industry?

“Life cycle analysis (LCA) is a methodology that allows us to assess the environmental impact of a product or service throughout its existence. From the extraction of raw materials to their final disposal.

It consists of mapping all the processes involved and calculating what is extracted from the environment and what is returned to it. From there, more specific studies can be derived, such as the carbon footprint (greenhouse gas emissions) or the water footprint (water use). It’s a very useful tool for making more informed decisions, both in design and communication.”

Q: Why did you decide to specialize in this?

“It seemed like a logical step. At Stämm I am in charge of the environmental area and I am part of the quality team, which is tasked with understanding and standardizing our processes. With that knowledge, we can improve what we do, optimize it, and make it more sustainable.

Quality and environmental care are not separate paths. At Stämm we have a strong culture of environmental commitment, and our products, by design, already consume fewer resources than traditional alternatives. By imitating natural processes, they are more efficient from their conception.”

Q: When is it most effective to apply sustainability criteria?

“Eco-design, designing with minimizing environmental impact in mind, is most effective in the early stages of product development. Implementing it late can be costly and limited.

That’s why this approach works for us now and in the future. It allows us to design products like our bioprocessor from a circular logic. It is also useful to show consumers and investors that our products are environmentally superior.”

Q: Is environmental responsibility part of the product?

“Absolutely. Quality, safety and environment are three inseparable pillars. A product that doesn’t do its job, that is not safe, or that harms the environment, is not a good product.

Unfortunately, quality or safety are often prioritized and environmental issues are left aside. But at Stämm this does not happen. Here we have a culture where everyone is involved and concerned about the impact of what we do”.

Q: What does the environmental perspective bring to product development?

“Something simple but key: all residue is wasted raw material. If I manufacture paper and waste two kilos of raw material per ream, I’m losing efficiency. The same applies to all processes. Making better use of resources reduces costs, waste and emissions.

In addition, knowing our processes allows us to comply with legal regulations. If we do not have mapped effluents, liquid, gaseous or solid, we may be generating environmental or safety risks without knowing it. Life cycle analysis would help us prevent that and avoid unforeseen events.”

Q: How are circular economy and life cycle analysis related?

“Circular economy proposes a model where waste is transformed into resources. In nature, there is no such thing as waste: everything finds its use in another process.

That is the principle that guides our work at Stämm. For example, we recycle plastic from our workshops and reuse it. We’re not in a perfect closed loop yet, but that’s where we want to go. And life cycle analysis is the compass that guides us on that path.”

Q: What are the advantages of modular design from an environmental perspective?

“Modular products, divided into parts that can be interchanged, have a distinct advantage: if one part breaks, you can replace just that part. That reduces waste and costs.

Our bioprocessors work this way. Each module is independent, allowing for simpler repairs and less environmental impact. Better for the planet and for the user.”

Q: Is there such a thing as zero environmental impact?

“It’s debatable, but no, not really. Every activity generates impact. Even living generates impact. But there are processes whose net impact can be very low, or even positive if energy is recovered. For example, planting a tree. The process of photosynthesis has a negative carbon footprint, as it captures CO2 from the air and fixes it structurally.

In the case of Stämm, our products already reduce water consumption, energy, space (footprint), and require less human intervention compared to current methods. The more their use scales up, the clearer the difference with other technologies will be.”

Q: Can efficiency become a problem in itself?

“Yes, it does. Take the example of refrigerators. Today they are more efficient, but because they are cheaper, there are many more of them. So the total impact is greater, even if each unit consumes less.

But I choose to be optimistic. I believe we are in a transition stage. First comes technological innovation, then optimization. Efficiency is growing and that is key.

Also, it’s not just about environmental impact, it’s also about human welfare. It seems to me that it is worth the cost of more people living better lives. And both can be balanced if we are smart about design.”

Q: Do you see progress from the engineering and regulatory sides?

“Yes. Consumers are becoming more aware, especially when they have their basic needs met.

There are also important regulatory developments. In Europe, for example, there is already a requirement to declare the environmental footprint of many products. And standards are becoming more demanding every year.

There is still a lot to do, but the trend is positive.”

Q: How do you envision the future of Stämm products?

“I think we are moving towards a model where the product is not only sold and where Stämm’s relationship with the customer does not end once the product is delivered, but comes with an integrated service: repair, maintenance, parts replacement and upgrade.

For example, if we bring out an improved version of a component, such as a sensor, you don’t need to change the whole equipment. Just that part. It’s like upgrading the cell phone camera without replacing the whole device. It’s cheaper, more sustainable and more convenient for everyone.

It’s a matter of aligning incentives from all sides: regulatory, economic and environmental.”

At Stämm, our vision is clear: to innovate by drawing inspiration from Nature, optimizing every process with critical thinking and environmental commitment.

Learn more about our mission and products at stamm.bio.

Introduction: A New Era of Bio-AI

The convergence of Artificial Intelligence (AI) and synthetic biology marks a bio-AI revolution across various life science domains, including medicine, agriculture, and industrial biotechnology.

This integration promises to cut development times and costs, enhance precision, and address complex challenges previously deemed intractable by traditional methods. The core premise is that AI, particularly machine learning (ML) and deep learning (DL), can effectively process and derive insights from the massive, complex data sets generated in biological research, accelerating the “Design-Build-Test-Learn” (DBTL) cycle inherent in synthetic biology.

AI, particularly ML and DL, is central to this revolution. AI systems are designed to simulate the intelligence of higher organisms and learn from experience. In this sense, ML and DL algorithms can pick up knowledge from data of various sources and become more advanced, intelligent, and self-aware artificial systems. They are adept at recognizing patterns from large volumes of information, providing more reliable results than manual analysis and predictions.

These tools can be applied in a huge number of cases and industries. Here we will focus first on drug discovery & safety, then on personalized medicine, and finally on synthetic biology in general, including bioengineering for biotechnology and biomanufacturing.

1. Drug Discovery Reimagined: AI and Patients-on-Chip

Drug discovery and safety is a complex, lengthy, and expensive process involved in discovering, developing, testing, and ensuring the suitability of a new treatment for a disease. It involves identifying biological targets, finding the appropriate compounds, and testing for effectiveness and safety.

The process is unbearably slow and expensive, costing on average over US$2.6 billion, and spanning 12–15 years. A staggering 89% of drug candidates that successfully pass animal testing fail in clinical trials, highlighting a significant clinical prediction challenge. — Bentwich.

The true revolution in the field of information technology (IT) has led to the production and storage of an enormous amount of data across various fields, including biosciences. Traditional model-driven methods struggle to uncover information, forecast data behavior, and comprehend complex data linkages from this big data, which includes data associated with clinical trials, drug trials, and experiments.

AI can accelerate the discovery of drug candidates by navigating a combinatorial space of more than 10⁶⁰ molecules that is simply too large for human beings to process effectively. — Lou.

It can identify novel antibiotics like Halicin in a fraction of the time that traditional methods require and optimize mRNA sequences for vaccines rapidly.

This high-throughput testing allows these algorithms to run thousands, later millions, of experiments on genetically diverse patients-on-a-chip to generate massive and highly predictive data on drug safety.

Unlike traditional methods, by using this kind of prediction, AI does the heavy lifting in determining whether a drug candidate will work safely in the human body.

AI significantly accelerates search and discovery in a previously computationally infeasible space, leading to faster identification of drug candidates. Some companies are leveraging AI for protein folding, while others focus on molecule design, target identification, and clinical trial optimization.

This approach to drug discovery has the potential to transform personalized medicine: a patient-centric approach that uses genetic, physiological, and environmental factors to tailor treatments and develop targeted drugs.

2. Personalized Medicine: AI, Patient-on-Chip & Multi-Omics for Tailored Treatments

Conventional 2D cell cultures offer limited biological relevance due to their inability to capture multi-cell complexity, tissue-like architecture, and physiological activity. Even advanced ‘omics data on its own struggles with these models, which poorly predict clinical safety.

But the current access to genetic information has changed how a patient’s individual profile can be understood. Doctors could use this information to make decisions on diagnsosis, prevention, and treatment, but also to choose medication and therapy.

For example, patient-on-chip technology (microfluidic devices that connect miniaturized, functional human organs) offers a more accurate in vitro model of the human body and could transform the current medical landscape.

By training ML on data from diverse patient-on-chip experiments, the system can predict not just general drug safety, but who is safe to use it, potentially salvaging failed drugs and optimizing clinical trials. We could predict which specific groups of patients (based on age, genetics, health conditions, etc.) are likely to use a drug safely.

But that’s not it for personalized medicine. AI allows researchers to manage challenging issues, including quantitative and predictive epidemiology, precision-based medicines, and host–pathogen interactions. Again, analyzing patient data and genetic makeup to create tailored treatment plans, not only minimizing treatment cost but also drugs side effects.

In this sense, AI facilitates customized medicine by considering individual genetic, metabolic, and physiological factors to tailor treatments, integrating multi-omics profiles to understand the complete picture. For example, AI-driven systems that generate personalized treatment plans often face challenges integrating with current healthcare providers, as they require time and availability from professionals that many systems can’t accommodate.

But it’s not just about personalized medicine; the impact is systemic: it’s a change at the level of synthetic biology in general, reaching all bio industries, biomanufacturing, bioengineering, biotechnology…

3. Synthetic Biology: Engineering Cells and Genomes for Industrial Biotechnology and Biomanufacturing

Synthetic biology (SynBio) is a multidisciplinary field that applies engineering principles to living systems. Its goal is to design and build new biological parts, devices, and systems or reprogram existing ones found in nature.

At its core, SynBio focuses on re-engineering organisms for useful purposes. This can involve creating biological modules, systems, and even machines, or redesigning natural pathways to perform new functions. To achieve reliable and predictable outcomes, SynBio uses systems engineering approaches, aiming to build biological functions that go beyond what evolution has already produced.

But there are still inefficiencies in bioengineering. Metabolic engineering, crucial for producing valuable molecules like biofuels and drugs, has been far from systematic, leading to very long development times.

AI and ML-powered synthetic biology employs an automated DBTL workflow for designing and constructing genetically reprogrammed organisms. AI is particularly well-suited for enhancing the “Learn” phase of this cycle, as it can predict biological system behavior and empower the Learn phase by analyzing high-throughput phenotyping data.

Cell bioengineering is leveraged to synthesize novel, valuable molecules such as renewable biofuels, monoclonal antibodies, or anticancer drugs. AI and ML can suggest modifications to the metabolic pathways of living systems to obtain the best possible outputs with the minimal inputs, improving industrial strains of microbial species to maximize the yield in the bio-based industrial setup.

These algorithms are accelerating bioengineering. Tools like the Automated Recommendation Tool (ART) leverage ML to guide synthetic biology efforts, speeding up the DBTL cycle for producing molecules like renewable biofuels (limonene) and flavor compounds (hoppy beer). ART provides quantitative prediction of the expected production, a systematic method that is automated, and uncertainty quantification for the predictions.

Finally, AI applications extend beyond medicine to agriculture for crop management, yield prediction, and optimizing resource use, as well as industrial biotechnology for enhancing bioproducts like biofuels.

Challenges and future directions

The adoption of Bio-AI is driven by traditional life science approaches, despite the immense potential, several challenges remain:

Current AI and ML are unable to determine whether drug candidates are clinically safe and efficacious. AI’s effectiveness in drug discovery is greatest for drugs with a medium level of chemical novelty and where the mechanism of impact on a disease is known; it is less helpful for entirely novel or incremental drugs. Also, problems with reproducibility in biological experiments remain a significant issue.

While AI thrives on large data sets, obtaining sufficient, high-quality, and diverse biological data remains a hurdle. The inherent variability in experiments and the multi-modal nature of biological data pose integration challenges.

Effective Bio-AI innovation requires a continuous synthesis of knowledge from both AI and medical experts, emphasizing the need for employees who possess a combination of AI skills and domain knowledge. This will benefit teams that know effective interdisciplinary collaboration.

The rapid advancement of Bio-AI raises critical ethical concerns and future societal challenges. Issues like data privacy and security, equitable access to technologies, and defining “humanness” in the context of advanced bioengineering require careful consideration and international cooperation.

Building the Next Chapter in Bio-AI

The future envisions self-driving laboratories that augment automated experimentation with AI, enabling autonomous experimentation and accelerating bioengineering. The ultimate goal for science AI is to move towards a quantitative feasibility score based on human evaluation and experiment feedback, integrating knowledge mining with real-world experiments.

The convergence of AI and synthetic biology is clearly reshaping the life sciences, from how we discover drugs to how we personalize treatments to how we engineer biology itself. Yet, as powerful as these technologies are, we’re still learning how to make them truly work with biology, not just around it. The challenges are real: noisy data, complex systems, and a need for better integration across disciplines.

At Stämm, we see these challenges as invitations to rethink how we build with biology. That’s why we’ve been developing MoNA, our Multi-omic Network Atlas, a structured way to connect the dots between genes, cells, conditions, and outcomes. It’s our approach to closing the loop between in silico and wet-lab, between data and decision.

With MoNA, we are offering a new lens. One that helps us (and our partners) choose the right cell lines, spot promising clones, and fine-tune the environments that cells need to thrive. It’s a foundational step toward something bigger: fully digital, adaptive, and sustainable bioproduction.

We’re still early in this journey. But with every dataset, every experiment, and every collaboration, we’re getting closer to a world where biology is understood, deeply and holistically.

And that’s a future we’re excited to build, together.

References

A Dixon, T., C Curach, N., & Pretorius, I. S. (2020). Bio‐informational futures: The convergence of artificial intelligence and synthetic biology. EMBO Reports, 21(3), e50036. https://doi.org/10.15252/embr.202050036

Bentwich, I. (2023). Pharma’s Bio-AI revolution. Drug Discovery Today, 28(5), 103515. https://doi.org/10.1016/j.drudis.2023.103515

Bhardwaj, A., Kishore, S., & Pandey, D. K. (2022). Artificial Intelligence in Biological Sciences. Life, 12(9), 1430. https://doi.org/10.3390/life12091430

Choudhury, M., Deans, A. J., Candland, D. R., & Deans, T. L. (2025). Advancing cell therapies with artificial intelligence and synthetic biology. Current Opinion in Biomedical Engineering, 34, 100580. https://doi.org/10.1016/j.cobme.2025.100580

Cizauskas, C., DeBenedictis, E., & Kelly, P. (2025). How the past is shaping the future of life science: The influence of automation and AI on biology. New Biotechnology, 88, 1–11. https://doi.org/10.1016/j.nbt.2025.03.004

Ezeako, E. C., Solomon, A. Y., Itam, Y. B., Ezike, T. C., Ogbonna, C. P., Amuzie, N. G., Aham, E. C., Aondover, C. D., Osuagwu, G. O., & Ozougwu, V. E. (2025). Prospects of synthetic biology in revolutionizing microbial synthesis and drug discovery. Life Research, 8(1), 6. https://doi.org/10.53388/LR20250006

Holzinger, A., Keiblinger, K., Holub, P., Zatloukal, K., & Müller, H. (2023). AI for life: Trends in artificial intelligence for biotechnology. New Biotechnology, 74, 16–24. https://doi.org/10.1016/j.nbt.2023.02.001

Iram, A., Dong, Y., & Ignea, C. (2024). Synthetic biology advances towards a bio-based society in the era of artificial intelligence. Current Opinion in Biotechnology, 87, 103143. https://doi.org/10.1016/j.copbio.2024.103143

Khan, A. S. (2025). Scope and impact of artificial intelligence and machine learning & deep learning in biology. Journal of Bacteriology & Mycology: Open Access, 13(2), 86–87. https://doi.org/10.15406/jbmoa.2025.13.00403

Lou, B., & Wu, L. (2021). AI on Drugs: Can Artificial Intelligence Accelerate Drug Development? Evidence from a Large-Scale Examination of Bio-Pharma Firms. MIS Quarterly, 45(3), 1451–1482. https://doi.org/10.25300/MISQ/2021/16565

Nesbeth, D. N., Zaikin, A., Saka, Y., Romano, M. C., Giuraniuc, C. V., Kanakov, O., & Laptyeva, T. (2016). Synthetic biology routes to bio-artificial intelligence. Essays in Biochemistry, 60(4), 381–391. https://doi.org/10.1042/EBC20160014

Radivojević, T., Costello, Z., Workman, K., & Garcia Martin, H. (2020). A machine learning Automated Recommendation Tool for synthetic biology. Nature Communications, 11(1), 4879. https://doi.org/10.1038/s41467-020-18008-4

Rice, A. J., Sword, T. T., Chengan, K., Mitchell, D. A., Mouncey, N. J., Moore, S. J., & Bailey, C. B. (2025). Cell-free synthetic biology for natural product biosynthesis and discovery. Chemical Society Reviews, 54(9), 4314–4352. https://doi.org/10.1039/D4CS01198H

Swapna, M., Viswanadhula, U. M., Aluvalu, R., Vardharajan, V., & Kotecha, K. (2022). Bio-Signals in Medical Applications and Challenges Using Artificial Intelligence. Journal of Sensor and Actuator Networks, 11(1), 17. https://doi.org/10.3390/jsan11010017

Xiao, Z., Pakrasi, H. B., Chen, Y., & Tang, Y. J. (2025). Network for knowledge Organization (NEKO): An AI knowledge mining workflow for synthetic biology research. Metabolic Engineering, 87, 60–67. https://doi.org/10.1016/j.ymben.2024.11.006

What does it mean to be a biomedical engineer?

“I think biomedical engineers are the industrial engineers of healthcare: we know a lot about many things, but often we don’t know that much about them in depth. A computer scientist, a doctor, an electronics engineer… they all know more than I do in their specific areas. I know a little about each one, which allows me to interact with different disciplines.

The nice thing about this career is that, even though you specialize later on, you can always change course. With the same basic training, two people can end up in completely different jobs. That versatility is valuable for curious people. I leaned toward computer science, did my thesis on machine learning, and that led to bigger decisions, like looking for a job in that field.”

What do you like about machine learning?

“I’ve always liked solving problems. Programming seems like a limitless tool for that. There’s a phrase I love: ‘If your code works, don’t touch it.’ It may not be perfect, but if it does the job, that’s already progress.

Beyond the code being functional, it has to be understood in the context you’re in. You work with other people, so it’s important that it not only works, but is also understandable.

As I was saying, with machine learning and programming in general, you don’t work alone. That’s why I started to value readability and documentation. Also, since much of my career took place during the pandemic, what I was able to experience most was the programming side. It was the most accessible from a virtual standpoint, and it ended up captivating me.

I like machine learning’s immediacy. You can create something that solves a real problem, and that’s very satisfying. For example, at Stämm, when I developed a model, Lynceus, to guarantee the quality of bubble-free bioreactors (the 3D cartdrige we produce), there was a back-and-forth with the operators who had been doing that task manually. We all wanted to automate it, but feedback was key. In that back-and-forth, they told me what they needed, and I was able to iterate quickly based on their feedback. That give-and-take is very rewarding. It has a pace that is sometimes difficult to experience in academia.”

And what do you like about bioengineering?

“What I love about my career is that it’s multidisciplinary. You can learn a little bit about programming, biology, physics, imaging… and that ‘little bit’ of each thing allows you to see the ‘big picture’. Understand how an image is captured, what it represents biologically, and what can be done with that information from a computational standpoint.

That global vision also allows you to talk to specialists in each area, and even if you’re not an expert, you can understand and collaborate with them. At Stämm, for example, there are lots of experts and passionate people. You learn a lot just by listening to them.”

What potential and limitations do you see in machine learning?

“Looking at the news, it’s obvious that AI and machine learning are on the rise.

But sometimes it seems like magic to the general public, when in reality it’s not. You need a lot of well-organized data to train a model. You spend a lot of time dealing with clients, some of whom come in with a lot of preconceived notions about machine learning. Some come with great ideas, but without the right data, nothing can be done. Without a structured and available data source, you can’t pull a model out of thin air.

Furthermore, the model learns from what you feed it. For example, imagine you are training a model to distinguish between living and dead cells. If you train it with images of one type of cell, it will not be able to recognize others. Or if all the living cells happened to have a purple spot, it will believe that anything with a similar spot is a living cell. There is a false illusion that you can throw anything into it and it will give you the right answer. But it doesn’t work that way.

There is also something important to consider: not all data is information; it needs to be processed. We have access to a lot of data every day, but the question is how to transform it so that it becomes valuable and effectively becomes information.

On the other hand, data does not always exist in an exploitable form. It needs to be organized in a particular way, and the model is extremely sensitive to the data you train it with.

Another important limitation is validation. If you generate, for example, a synthetic cell, how do you know that it makes biological sense? You need to validate it, see if it has a coherent transcription, if the basic pathways are active. That’s the difference between having a good statistical model and a biologically useful one.

I think that’s where the breadth of my training in so many different subjects helps me. For example, I know that whether a cell is alive or dead, it interacts with radiation and light differently. And that lets me know that when you record the image, it will look different. That’s where I use the more biological part of interpreting that data combined with the computer science side of seeing what I do with that information.

That’s the great thing about being able to talk to experts in different fields and understand what they’re saying. I don’t need to know everything, but I do need to have a solid foundation that allows me to have a conversation with other specialists.”

What role does documentation play?

“I like documenting. I like writing. Writing helps me understand what I do, and it also allows others to understand and use my work. I think it’s important because we often work in teams, and the code must be clear and shareable.

Documentation also helps to ground complex processes. Machine learning involves many somewhat ‘artistic’ decisions. Sometimes you don’t know why something works, but it works. I try to unravel that process by writing, diagramming… whatever helps me understand.

Working in a team forces you to adopt these practices as well. Code is very personal, and you have to leave things precise and understandable for those who come after you or work with you.”

What risks do you see in machine learning?

“Sometimes it seems like magic, but in certain ways it’s still a black box. It’s like being a data alchemist: you mix things together and suddenly you get something useful, sometimes without being entirely sure what happened.

But when that applies to health, you have to be very careful. A model that predicts a disease should have the same level of validation as a laboratory test or medical equipment. We must ensure that this is also the case today and that there are clear and up-to-date regulations in this regard.

It feels a bit like machine learning is like the internet in the 90s: everyone is trying things out, with a lot of potential but at the same time that frontier feeling, like the wild west. It’s great for creativity and its mass appeal, but we should make sure it doesn’t get out of control.

When I got into the world of machine learning, my biggest question was, well, is this how it is and we all accept that it is? Or can I reach a level of experience and understanding and be sufficiently trained to break that barrier and understand why things are working this way?

The answer is a bit yes and a bit no. That is, you will be able to understand more as you understand more of the mathematics and statistics behind the models, but there comes a point where no one knows what is going on anymore.

For example, if you take ChatGPT: there are a lot of people who understand the theoretical basis of why it works the way it does in general, but because it is a model based on neural networks, it has certain aspects that are very opaque to decipher.

It’s something that machine learning has. You design a model, you say it’s going to work and this is going to go in here and give me this output, and out of nowhere something totally random happens. And you say, why are you doing this?! I designed you to do this whole process and give me this output.

In language models, this thing called “hallucinations” emerged. Basically, the language model starts hallucinating and says things that have nothing to do with the information it read. Then you wonder, it even seems like, I don’t know, a living being. I gave you all this information, and yet you’re making up something else. What it says doesn’t make sense, but the question is, how did you come up with that connection?”

What future do you want for machine learning?

“I would love to see certain processes standardized, with clear regulations on what can be done and how. We need to be aware that we are affecting people’s health and act accordingly.

I’m also very interested in preventive diagnosis. Models that can predict diseases before they appear, such as some that are already used with mammograms. I think that’s the greatest potential of machine learning in healthcare.

I believe it really has applications that can save lives if we use it responsibly. We must never forget that this is patient health data and it must be treated with discretion.

This is the area that would motivate me most to tackle and innovate: solving problems even before they arise.”

“Those who watched the first starved, sometimes comatose, patients receive insulin and return to life saw one of the genuine miracles of modern medicine.”
 — Professor Michael Bliss.

Before insulin, a Type 1 diabetes diagnosis was a slow goodbye.

Children wasted away. Adults faded into comas. Doctors tried starvation diets to delay the inevitable.

Then, in the early 1920s, something extraordinary happened: insulin turned the tide.

And biotech, as we know it today, was born.

A 2AM idea that rewrote medical history

In 1920, Dr. Frederick Banting, a young Canadian surgeon, scribbled a strange note in his journal:

“Ligate pancreatic ducts of dog. Keep dogs alive till acini degenerate. Try to isolate the internal secretion to relieve glycosuria.”

That moment kicked off what would become one of the most dramatic transformations in medical history.

What followed was a summer of relentless experimentation. Banting and medical student Charles Best worked with minimal funding, stray dogs from the streets of Toronto, and a lot of guts.

They succeeded where plenty of scientists had failed.

In January 1922, they injected a purified pancreatic extract, prepared by biochemist James Collip, into 14-year-old Leonard Thompson. The first attempt failed. The second worked. Leonard came back from the brink.

“They were present at the closest approach to the resurrection of the body that our secular society can achieve,” wrote historian Michael Bliss​.

Why they succeeded (when so many hadn’t)

Banting and Best weren’t first. Researchers like Nicolae Paulescu, Zülzer, and Kleiner had all tested pancreatic extracts before 1921 [source]. But none could purify the substance enough to make it safe and effective for humans.

What changed?

• Access to better tools: New glucose tests gave faster, more precise feedback.
• A lab to work in: J.J.R. Macleod provided a well-equipped, university-backed lab.
• Biochemist backup: James Collip joined to purify the extract for human use.

And maybe most importantly:

“Banting wasn’t just curious — he was determined to isolate something useful to treat diabetes.” — Lewis & Brubaker, JCI.

From pig pancreases to genetically modified E. coli

Early insulin came from animal pancreases: millions of pigs and cows, processed in industrial-scale facilities. It was lifesaving, but unsustainable.

Then came the next biotech revolution.

In 1978, Genentech and Eli Lilly used recombinant DNA to engineer human insulin in E. coli bacteria. This was:

• The first genetically engineered drug approved for humans
• A turning point in biomanufacturing
• A model for future biologics, from growth factors to monoclonal antibodies

“This marked the birth of modern biosynthesis and biomanufacturing — insulin showed it could be done.” — Mayer et al., Biopolymers.

“Insulin’s ability to restore health is so dramatic that its clinical pharmacology was initially described in biblical-like terms, such as ‘the raising of the dead,’”— Mayer et al., Biopolymers.

100 years later: A triumph with unfinished business

Accessibility remained a crisis, even after this breaktrough.

Di Bartolo and Eckel, both of whom also live with Type 1 diabetes​, recount a past where syringes needed boiling, travel required portable refrigerators, and every meal was pre-planned.

The 1985 arrival of insulin pens changed everything, empowering patients with flexible, discreet care. Later came fast-acting analogs, wearable pumps, and algorithm-driven dosing.

“Insulin was once seen as a barrier to freedom. Today, with innovation, it’s a tool for autonomy.” — Di Bartolo & Eckel.

Insulin today is better than ever, it has:

• Fast-acting analogs
• Pen injectors and pumps
• Closed-loop systems
• Algorithm-driven dosing

Despite all progress, millions still lack access to affordable insulin. What started as a freely shared discovery, Banting sold the patent for $1, has become a multi-billion-dollar industry with complex IP landscapes and high costs for patients.

A molecule discovered as a gift to humanity has become, for some, a symbol of inequity.

The therapeutic burden remains high. As Mayer and colleagues note, insulin has a narrow therapeutic index, is injected, and has fragile stability: not exactly an ideal drug from a delivery perspective​.

Why the insulin story still inspires us at Stämm

The insulin story isn’t just biotech history, it’s a playbook.

It didn’t just save lives. It taught us that biology could be scalable.

It’s a story of radical collaboration across disciplines. Of humans daring to design with nature. And of a therapeutic built not just in a lab, but in the resilience of patients, many of whom helped shape how insulin evolved.

It reminds us that biotech doesn’t scale by accident. It takes:

• Curiosity, courage, and cross-disciplinary grit
• Tools that match biology’s complexity
• Systems designed for flexibility, access, and sustainability

At Stämm, we’re applying those lessons to the next generation of biologics. We’re building fluidic systems that grow biologics the way nature would, decentralized, adaptive, and efficient. Because it’s not just what we make, it’s how we make it.

TL;DR: Why insulin still matters

• It made history: Insulin saved lives and launched biotech.
• It shaped technology: rDNA insulin proved complex biology could be scaled.
• It challenged access: Even today, insulin isn’t universally affordable.
• It inspires us: We need biomanufacturing that’s faster, fairer, and nature-aligned.

Want to see how we’re rethinking how biologics are grown?
Follow us here on Medium. Join the conversation. Let’s build the future, together.

“In deeptech, you must have a vision of the future inside your company.” — Yuyo Llamazares.

Today, it feels easier to imagine the end of the world than to envision a radically better one. As Fredric Jameson famously put it, dystopia has become more imaginable than systemic change. Mark Fisher echoed the same concern: “Is there no alternative?”

At Stämm, we believe there is.

When we looked at the biotech industry, and the wider world, we saw two dominant futures on offer:

• Accelerationist techno-optimism: Technology will somehow save us if we just keep moving faster.
• Tribal primitivism: Longing for an imaginary past without technology.

Both of these narratives shape our media and collective imagination.

Think of Blade Runner: its 1982 vision of corporate-dominated, anti-human cities still defines much of our pop-cultural future. Even Blade Runner 2049 retraces the arc of ecological collapse first mapped by Philip K. Dick’s book, Do Androids Dream of Electric Sheep?.

If we don’t consciously propose better futures, we risk defaulting to these bleak ones.

Inspired by Nature

The story goes that Yuyo, one of our founders, was in a dilemma: the yeast for their beer venture wasn’t growing the way he wanted it to.

One day outdoors, in a recently mowed garden with the smell of fresh-cut grass, he found himself looking at a leaf backlit by the sun. Anyone who has done it knows about those veins that deliver and bring nutrients for the survival of every cell.

“What if we could replicate this system?” he thought. This perhaps innocent-looking question is what brought us here: bubble-free bioreactors where we can grow cells mimicking nature’s vascular systems.

From Competition to Collaboration

Nature teaches us that survival doesn’t depend solely on competition, it depends on collaboration.

In ecosystems, species thrive together in complex, symbiotic networks.

This principle must guide biotech innovation too.

Instead of building isolated technologies for narrow gains, we must build ecosystems: partnerships between academia, industry, educational institutions, and startups.

At Stämm, we committed to a long-cycle vision:

Decentralize cell production and transform the global manufacturing of pharmaceuticals, biomaterials, and biomass.

A purely technical solution might have been easier. A short-term business might have been simpler. But deeptech works on a long life cycle, it requires patience, purpose, and community.

Solarpunk vs Cyberpunk: Choosing Our Future

The dominant stories we tell about the future matter.

That’s why we consciously reject the cyberpunk dystopia, however compelling its literary and cinematic worlds may be.

Instead, we draw inspiration from Solarpunk: a genre of climate fiction that envisions futures based on:

• Sustainability
• Creativity
• Regenerative technology
• Local autonomy

In this emerging narrative, solutions like solar-energy stained glass, zero-emission travel, and high-altitude wind farms aren’t distant dreams: they’re becoming realities.

We see a future where clean energy and distributed systems replace centralized, extractive models.

In an incremental but significant way, we hope our bioprocessor technology will contribute to this post-scarcity, community-driven world, a world more akin to the imaginative universes of Studio Ghibli than the cyberpunk cityscapes of neon decay.

Local Roots, Global Impact

Just as NASA discovered that Saharan dust feeds the Amazon rainforest, we believe that innovation from seemingly remote regions, like Argentina, can nourish global ecosystems.

Argentina has a long tradition of speculative futures, from bleak ones like The Eternaut (1957–59), a classic of Argentine sci-fi, to Michel Nieva’s recent gaucho-punk fiction published by Caja Negra, and Borges’ remote future presented on Utopia of a Tired Man, inspired by Ray Bradbury.

We are not naive: we know technology can reinforce inequities if we aren’t vigilant.

We understand that tools shape societies, just as narratives shape possibilities.

Still, we dream and work, toward a better future, rooted in resilience, creativity, and collective stewardship.

With the support of our partners: GridX, Singularity University, Draper Cygnus, IndieBio, BioArk, and others, we move forward, aware that building better futures is not just a market opportunity.

It’s a responsibility.

What future will you help create? Learn more at stamm.bio.

References

Nasa satellite reveals how much saharan dust feeds amazon’s plants

Capitalist realism, is there no alternative?, Mark Fisher: https://cajanegraeditora.com.ar/libros/realismo-capitalista/

Nuestra propia fábrica de yogures, El Gato y la Caja: https://elgatoylacaja.com/newsletter_clima/te-amo-solarpunk-pero-vamos-a-necesitar-mas-de-vos

Ficciones gauchopunks, Michel Nieva

Utopia of a tired man, Jorge Luis Borges

I always wanted to help improve people’s quality of life, whether by providing information, collaborating in clinical trials or finding solutions to health problems.

1. Could you tell us about your training and specialization?

I am a biologist, specialized in molecular biology. My higher education was focused on human diseases.

During my PhD, I worked on the study of psychiatric diseases associated with stress, such as major depressive disorder and bipolar disorder. I focused on the dysregulation of the hypothalamic-pituitary-adrenal system and its relationship with the onset of these pathologies.

My research analyzed the effects of antidepressants beyond the traditional neuronal impact, exploring non-canonical pathways that, properly used, can be beneficial for certain pathologies.

I did part of my PhD in Argentina and another part in a neuropsychiatric center in Germany, thanks to a collaboration between both institutes. Then, I did a postdoctorate in the same specialization, deepening one of the lines of research that remained open during my thesis.

2. What motivated you to choose this career path?

I always wanted to help improve people’s quality of life, whether by providing information, collaborating in clinical trials or finding solutions to health problems.

I contemplated studying medicine, but felt that I would not adapt well to the constant contact with death and the emotionally complex situations that physicians face.

In addition, I was more attracted to biology and chemistry. I think it is important to know yourself and where you can contribute your best.

3. During the pandemic, you collaborated in COVID-19 diagnostic tasks. How was that experience?

Yes, I worked as a volunteer at a hospital doing PCR testing. I felt that, having received a public education in Argentina, the least I could do was to contribute in a time of need.

It was interesting to work with physicians, as our backgrounds are very different. They are trained to solve emergencies quickly and efficiently, while in molecular biology we focus on a meticulous and controlled approach.

That said, this interdisciplinary collaboration was enriching and reaffirmed how much progress can be made when different perspectives are combined.

4. What led you to make the leap from academic research to biotechnology?

Although I love research, I increasingly felt the need to do something more tangible, something that I could see applied to a real-world problem. In academia, a lot of research never makes it to clinical trials or concrete applications. I became interested in biotechnology and medical trials, looking for a place where my work would have a more direct impact.

I came to Stämm through GridX, and I was attracted to its mission to democratize biomanufacturing. I believe that in many areas there is a gap between scientific knowledge and its practical application, and Stämm seeks to fill that gap, bringing solutions to more people and reducing the costs of personalized medicine.

5. How do you see the evolution towards personalized medicine and what is the role of biotechnology in this change?

It is becoming increasingly clear that we are not all the same and that responses to treatments vary according to age, sex, gene pool and other factors.

Many of the vulnerable groups (children, the elderly, pregnant women, people with pre-existing diseases, specific ethnicities, etc.) are not part of the majority of studies prior to the release of a drug to the market, and safety and efficacy in these populations are unknown.

The problem is that tailoring treatments to each individual is very costly due to current production methods. Therefore, the challenge is to find ways to reduce these costs so that personalized medicine becomes accessible to more people.

What I liked about Stämm’s speech is precisely this objective: to make it possible for the production of biomolecules, which today only takes place in industrial plants, to be carried out in hospitals or even in smaller spaces, adapting the interventions to each patient.

6. How do technology and interdisciplinary work at Stämm contribute to this goal?

One of the most valuable aspects of Stämm is its interdisciplinary approach. Working with experts from different areas (such as microfluidics, chemistry and simulations) broadens perspectives and generates innovative solutions. The collaboration has allowed us to develop a bioreactor with a morphology and resin that did not exist before.

From biology, we are used to controlled and meticulous processes, but industry times are faster. So we have to find a balance between getting fast results and collecting as much quality information as possible. The creativity of my colleagues has helped me to apply new ideas in my field, and I believe that constant interaction between different disciplines is key to solving complex problems.

7. What are the main challenges you are currently facing at Stämm?

As we look for a paradigm shift, we are continuously facing new challenges. For example, we wanted a microfluidic 3D chip, but the resins on the market were not biocompatible for cell cultures, so we developed our own resin. Then, the resolution of commercial printers was not sufficient, so we created our own printer. Each step brings new limitations that we have to overcome.

We also worked on developing cell lines capable of withstanding industrial conditions, which is essential to ensure the viability of the system in real applications. Solving these problems requires creativity and perseverance, but that is precisely what makes this work so exciting.

8. How do you envision the future if you achieve your goals?

I like to think of a future where our bioreactor, the bioprocessor, will be present in hospitals, institutes and companies, allowing us to produce biomolecules continuously and at lower cost. Hospitals could offer personalized therapies, growing patient-specific cells in small reactors that optimize time and resources.

In addition, reducing production costs could make biologic drugs more accessible, benefiting people with rare diseases that are currently unprofitable for the pharmaceutical industry. I believe that biotechnology has the potential to transform the way we approach healthcare, and my motivation is to help make that future a reality.

9. Finally, what inspires you to continue working with such dedication?

What excites me most is the thought of what people will be able to achieve with the tools we are developing. Just as the personal computer and the Internet opened up a world of possibilities, our goal is to make biomanufacturing available to more people so that they can create solutions that we can’t even imagine today. That future is what drives me to do my best every day.

Want to know more? Read our other articles on Medium or delve deeper at the future of biomanufacturing at stamm.bio

Q: Let’s start with the basics — what’s the difference between biological data analysis and bioinformatics?

Ezequiel Pulicari: The two are closely related. Bioinformatics uses computational tools to derive biological insights, so in that sense, they overlap. However, bioinformatics often refers specifically to working with biological sequences, particularly in the field of “omics” — like genomics, transcriptomics, metabolomics, and proteomics.

In my case, I focus on transcriptomics, which involves analyzing the RNA transcripts of a cell. Transcripts are the molecules that carry genetic instructions from DNA to make proteins. By sequencing and quantifying these transcripts, we can uncover how genes are expressed in different cells and conditions.

Q: How does transcriptomics help us understand cells better?

The genome — the complete set of DNA — is the same in all cells of an organism, but not all genes are active at all times. Cells differ because they express different genes in varying amounts. This difference in gene expression creates diverse cell types and phenotypes.

Phenotypes can range from observable traits, like a cell’s shape, to molecular characteristics, like the expression of specific genes. By analyzing transcripts, we associate certain gene expression patterns with particular phenotypes. For instance, one cell might be a neuron while another is a skin cell, all because of differences in gene expression.

Q: What does a typical transcriptomics dataset look like?

A human cell has approximately 150,000 different transcripts, corresponding to about 20,000 coding genes. When we analyze samples, we generate large tables where each row represents a gene and each column represents a sample or condition. At Stämm, our datasets can include up to 21,000 genes across tens of thousands of samples — massive amounts of data.

These datasets are far too complex for manual analysis. Instead, we use computational tools to identify patterns, such as which genes are expressed differently between conditions. This is known as differential gene expression analysis, which helps us focus on meaningful differences by filtering out “housekeeping genes” that are consistently expressed across all cells.

Q: How do you move from identifying differences to drawing biological insights?

After identifying differentially expressed genes, we enrich the results by associating them with biological pathways — networks of genes and proteins involved in specific processes, like metabolism or immune responses. This helps us see not just individual gene differences but how entire systems are behaving differently between conditions.

For example, we might find that 2,000 genes are differentially expressed between two cell types. By mapping these to pathways, we can identify processes that are more active in one cell type, such as stress responses or growth regulation.

Q: What are the practical applications of this work?

For instance, we might study differences between pluripotent cells (cells that can become any type of cell) and differentiated cells to optimize reprogramming processes.

We can also compare cells under different conditions, such as varying temperatures or nutrient levels, or even analyze patient samples. In one scenario, we might compare transcriptomes of insulin-sensitive versus insulin-resistant individuals to identify molecular differences — useful insights for pharmaceutical research.

Q: Beyond the lab, how is this relevant to drug discovery?

Transcriptomics is crucial for comparing how groups of cells respond to treatments, including drugs. For instance, we can analyze changes in gene expression caused by a specific drug to better understand its effects or identify potential side effects.

Another emerging area is protein modeling, such as predicting how a drug interacts with specific protein targets. While this isn’t my focus, tools like AlphaFold are revolutionizing this field by predicting protein structures with remarkable accuracy.

Q: You transitioned from biology to bioinformatics. How did that journey unfold?

During my biology studies, I had a knack for programming. While finishing my degree, I worked in a lab where I was assigned to analyze data from soil samples under different agricultural treatments. To speed up my work, I learned Python intensively and discovered how powerful computational tools could be.

When I joined Stämm, I continued to expand my skillset. The field constantly evolves, so I’m always learning plenty of new techniques and tools.

Q: Speaking of Transomics, what role do you play at Stämm?

My primary role is as a bridge between biology, statistics, and machine learning. I handle transcriptomics data analysis and collaborate closely with our machine-learning team to develop predictive models. This ensures that we focus on meaningful analyses and avoid wasting time on unproductive paths.

Q: What excites you most about the future of bioinformatics?

Two areas stand out. First, I’m excited about the potential of personalized medicine. Improved sequencing technologies could give us better insights into individual genetic predispositions and health risks.

Second, I’d love to see advances in cellular reprogramming. Imagine taking cells from one part of the body, reprogramming them, and turning them into another type of cell — like creating heart cells for a transplant. While we’re far from this being routine, progress in this area has been astonishing over the last two decades.

Q: How do you see your work contributing to these possibilities?

One vision for Transomics is to use machine learning to predict cellular behavior. For example, if we had comprehensive transcriptomic data for each cell, an algorithm could theoretically guide us on how to transform one into the other.

While we’re not there yet, we’re laying the groundwork by analyzing transcriptomic data and refining our models. It’s a step toward unlocking the potential of regenerative medicine.

Q: Any parting thoughts for aspiring bioinformaticians?

Understanding statistics and experimental design is critical, not just for bioinformatics but for science and life in general. It’s about systematically testing hypotheses — whether in a lab or while perfecting a recipe at home.

For anyone entering the field, I’d say this: stay curious, be open to learning, and don’t be afraid to dive into programming. The tools you master today could open doors to incredible discoveries tomorrow.

Q: Can you tell us about your background and how you joined Stämm?

I’m Daniela Ortíz, a molecular biologist at Stämm. I completed my PhD at the University of Buenos Aires (UBA), focusing on basic science. My research centered on the molecular biology of Drosophila (fruit flies), a common model organism in genetics, to understand the regulatory machinery of genes.

After an internship abroad, I joined Stämm, where I became part of the Tissues team. My role involves culturing stem cells, which can differentiate into any cell type.

Q: What does your work at Stämm involve?

At Tissues, we grow and multiply cells, then differentiate them into specific cell types and test their behavior in our proprietary bioreactors — devices designed for efficient cell cultivation.

We compare our bioreactors’ performance with commercially available wells, widely used in labs today, to assess how well our devices meet researchers’ needs.

Additionally, I contribute to Transomics, a platform designed to streamline lab work for biologists by simulating protocols and optimizing resources.

Q: How does Transomics simplify lab work?

In Tissues, differentiating cells requires testing various protocols from the literature. These protocols differ depending on available budgets, infrastructure, and reagents, making the process resource-intensive.

Transomics helps by simulating these protocols based on what researchers have on hand, saving time and costs. For example, instead of running multiple physical tests, a biologist could identify the best conditions through simulation, reducing months of work to hours.

Q: What broader applications do your tools support?

Transomics isn’t limited to streamlining lab protocols. It also supports cell therapy, gene therapy, and biomanufacturing. My primary focus is cell therapy, which relies on cells that can differentiate into any other type.

My role involves bridging lab work and computational models. This includes organizing and optimizing experimental setups to improve outcomes.

Q: How does computational biology enhance these processes?

Most biologists excel in designing experiments, but analyzing data often requires computational tools that not everyone is trained to use. This is where bioinformatics comes in, enabling the creation of algorithms, identification of key parameters, and execution of statistical tests.

Transomics takes this a step further by acting as a copilot for biologists. It assists with decision-making by suggesting protocols, connecting researchers with similar experimental goals, and even recommending relevant literature.

Q: How do you ensure Transomics is user-friendly?

My role involves providing biological insights to the development team while acting as a user myself. This ensures the platform remains practical and accessible for end-users.

For instance, we’re building a recommendation system that suggests potential assays, useful papers, and connections with researchers who have conducted similar experiments. This simplifies lab work and improves efficiency.

Q: What data feeds into Transomics, and how does it support other products?

Currently, Transomics uses public datasets and we’re supplementing it with our own experimental results. This additional data enhances the platform’s accuracy and benefits other products like the Human Tissue Platform (HTP), for scalable cell cultivation.

The data also feeds into a feedback loop: experiments generate data that trains the model, improving its ability to simulate future experiments with greater precision.

Q: Does your current work connect to your previous research?

In my PhD, I studied enhancers, regulatory elements that control gene expression. At Stämm, we focus on understanding gene expression in different conditions to better optimize cell behavior.

For example, studying gene expression requires analyzing RNA (the intermediate molecule between DNA and proteins). Using techniques like RNAseq, we can isolate and quantify the total RNA in a cell and thereby determine which genes are active under specific conditions.

Q: What is the potential of Transomics in the future?

The potential is enormous. Imagine being able to simulate experiments and eliminate costly, time-consuming lab work. Tasks that take months could yield results in hours.

This would free up time for biologists to design better experiments and focus on interpreting data, improving productivity and decision-making.

Q: How would your daily work change with a fully developed Transomics?

It would feel very different. I’m used to hands-on lab work, but with Transomics, I’d have the tools to analyze data independently.

For example, I could simulate experiments, analyze results, and plan new protocols without needing external support. It would give me greater control over my workflow.

Q: What other applications could this technology have?

The possibilities are endless. By analyzing different omics layers (DNA, RNA, proteins, and more), we can understand a cell’s state at multiple levels, answering virtually any biological question.

This approach isn’t limited to applied technologies like cell therapy — it brings us closer to understanding the phenotype (the way a cell “looks”) of any cell in an organism at any time, paving the way for discoveries in multiples areas of the biology such as physiology, genetic, evolution, ecology, botany, etc…

Q: What challenges do you foresee?

The biggest challenge is gathering enough data to train the models. Large language models had the benefit of being able to use written language: data that’s widely available to train the models. Biology assays are not as easy to find, interpret, and incorporate. Each cell type requires extensive datasets, and expanding to multiple cell types and organisms will only increase this need.

However, once a framework is developed for one cell type, it can be adapted for others, reducing costs and time incrementally.

Q: How has your role improved team collaboration?

We’ve implemented workshops to improve communication across the team. While we’re not all experts in each other’s fields, we now have a better understanding of how our roles intersect.

By explaining molecular biology concepts, I’ve helped align the team’s work with the needs of biologists, ensuring our tools are practical and effective.

Q: How do you bridge knowledge gaps for non-biologists in the team?

For team members without a biology background, I provide foundational workshops. These sessions cover basic concepts like the central dogma of molecular biology (which explains how genetic information flows from DNA to proteins, and the mechanisms involved in those processes) and the emergence and study of the existing omics.

For example, when we discuss omics fields:

• Genomics focuses on DNA and genes. DNA holds genetic instructions.
• Transcriptomics focuses on RNA (gene activity). RNA transmits these instructions.
• Proteomics focuses on proteins (functional molecules in the cell). Proteins carry out the cell’s functions, from building structures to accelerating chemical reactions as enzymes.

Understanding these basics helps the team design better tools for researchers and grasp more complex concepts like gene expression and its role in research.

Q: Can you share some closing thoughts?

Molecular biology is the study of life — it can be applied to anything. At Stämm, we focus on areas like cell and gene therapy, clinical assays, and transomics, but the possibilities are limitless. Tools like Transomics not only save time but also enable biologists to push boundaries, transforming both research and biomanufacturing.

To learn more about our Transomics platform visit our site: www.stamm.bio