Name, Position and Company
Fernando Oliver, CEO & Founder at Mafer AI.
Tell us about the company you work for. What is its story?
Mafer AI was founded in Barcelona by two childhood friends: Marc Montalbo (CTO, mathematician and Machine Learning researcher) and myself. We’ve known each other since we were five years old, and ever since our teenage years we knew we wanted to build something meaningful together. Each of us prepared for that moment in our own way: Marc from a deeply technical perspective, and me from a business-oriented one.
We found our opportunity in the specialty chemicals industry. Sectors such as fragrances, flavours and cosmetics compete through innovation and have accumulated decades of experimental knowledge, yet they still operate using technology that is ten to fifteen years old. Highly valuable information is scattered across Excel spreadsheets, laboratory notebooks, analytical instrument outputs and the expertise of just a handful of specialists. At Mafer, we are building the first data and artificial intelligence platform capable of capturing, normalising and structuring all this chemical knowledge. Our platform integrates directly with laboratory equipment (GC-MS, LC-MS, spectrometers), ERP systems and existing workflows, transforming fragmented information into structured intelligence ready to train proprietary AI models for each company.
We started working full-time on the company in October 2025, and growth has been remarkably fast. Today we are a team of twelve highly specialised professionals, including PhDs in chromatography and mathematics, engineers and AI/ML researchers. We work with leading fragrance, flavour and consumer goods companies across Spain and Europe, have successfully closed an investment round with top-tier funds and investors, and are part of the AI Factory at the Barcelona Supercomputing Center. Our ambition is to build, from Barcelona, a global company: the operating system used by R&D teams across formulation industries.
What does your day-to-day role involve?
My role sits at the intersection of customers, product and strategy. At Mafer, we often say that you either build or you sell—and I sell. I spend much of my week with customers, often in their laboratories and production facilities, understanding their chromatographic analysis, formulation and regulatory processes, and translating their operational bottlenecks into concrete solutions together with our technical team. I also work closely with executive teams, helping them transform their organisations into AI-native companies.
The rest of my time is dedicated to building the company itself: recruiting exceptional talent, maintaining relationships with investors and partners such as the Barcelona Supercomputing Center (BSC), and ensuring that every deployment delivers measurable business value. We operate with a Forward Deployed Engineers model, where the entire team works closely with customers. My ultimate responsibility is making sure that what we promise in a meeting becomes a fully operational product running on real customer data within a matter of weeks.
What makes Mafer AI different?
I would highlight three key factors, all closely connected:
True vertical specialization: We are AI-native by design, but we also understand chromatograms, raw materials, regulatory constraints and the logic behind formulation. Our team combines expertise in chemistry, mathematics and engineering, allowing us to handle chemical data with the level of scientific rigour it requires while applying the most suitable AI model architectures.
Seamless integration: Today, AI allows software to adapt to people—not the other way around. Our platform connects directly to raw laboratory data, ERP systems, LIMS and existing business workflows without forcing anyone to change the way they work. Within just a few weeks, customers begin to see the real impact AI can have on their R&D teams.
Customer data remains the customer’s intellectual property: Every company has its own private environment, and its data is used exclusively to train its own proprietary models. Data is never shared or reused across customers. We believe that in the future every company will have its own AI models, and their performance will depend on how well their data has been structured, curated and parameterised.
What challenges do you expect to face over the coming years?
The first challenge is scaling our delivery capabilities without compromising quality. Today, demand exceeds what we can currently serve, so our priority is to grow our team, processes and product while maintaining the technical excellence and close customer relationships that have brought us this far.
The second challenge is product evolution: moving from individual modules that solve specific bottlenecks to a comprehensive data platform and, ultimately, to foundation models capable of deeply understanding the chemical domain.
The third challenge is expansion. We aim to grow from fragrances and flavours—arguably the most technically demanding vertical—into cosmetics, food, home care and personal care. These industries all share the same foundations: molecular data, formulation logic and expert decision-making. Our goal is to bridge the long-standing gap between what scientific research already makes possible and what industry actually applies in its day-to-day operations.
What does being part of Beauty Cluster bring to Mafer AI as a company, and to you personally as a professional?
For Mafer, Beauty Cluster provides direct access to the ecosystem where we believe we can create the greatest impact. Fragrances, cosmetics and personal care are core verticals for us, and the cluster enables us to connect with brands, manufacturers, suppliers and laboratories that face the very data challenges we are solving every day. It is the ideal environment to listen to the industry’s needs, validate new solutions and build collaborations that accelerate technological adoption across the sector.
On a personal level, it provides valuable context and an outstanding professional network. Conversations with people who have spent decades in the industry help me better understand the entire value chain, from raw materials to the end consumer. For someone building technology for this sector, that continuous learning is invaluable.
Let’s talk about you. How did you end up in the beauty industry?
My path has been somewhat unconventional. I come from the world of investment and corporate finance (venture capital and M&A), not from the laboratory. It was while analysing industrial companies that I discovered a fascinating paradox: industries such as perfumery and cosmetics are incredibly innovative when it comes to products, yet very few AI startups were dedicating themselves to building solutions that matched that level of innovation.
Our entry point came through the Barcelona Perfumery Congress. We were fascinated to learn that a single fragrance may contain between 150 and 200 compounds, with an enormous amount of analytical work behind every formulation. Once Marc and I realised that all this complexity could be structured into data and transformed into AI models, we knew we had found the problem we wanted to dedicate ourselves to solving.
Since then, I have spent hundreds of hours in laboratories alongside perfumers, flavourists and analytical chemists, and we are working to establish ourselves within one of Spain’s most historic and technically sophisticated industries.
Speaking of beauty, how would you define it?
From my perspective, beauty is complexity perceived as simplicity. Behind a fragrance that moves you emotionally are hundreds of molecules in balance, years of accumulated knowledge and an extraordinary amount of scientific rigour. Yet the experience itself feels immediate and effortless.
That contrast between the sophistication behind the scenes and the simplicity of the final experience is, to me, the best definition of beauty. In many ways, it is also what we strive to achieve at Mafer: hiding all the complexity beneath the surface so that the expert’s experience remains simple and intuitive.
Do you have any hobbies, hidden passions or superpowers you’d like to share?
My superpower—if it can be called that—is applied curiosity. I enjoy diving into subjects I know nothing about, asking countless questions and, within a few weeks, understanding them well enough to have meaningful conversations with people who have spent decades working in that field. That happens with chemistry, but also with almost anything that sparks my interest.
Outside of work, what recharges me the most is spending time with my family and enjoying long meals with lifelong friends. After all, Mafer itself was born from a friendship that began more than twenty years ago.
What’s the last book you’ve read or the last film you’ve watched?
The last thing I read wasn’t exactly a book, but rather Pope Francis’ encyclical Magnifica Humanitas on artificial intelligence. It caught my attention because if the Pope is dedicating a profound reflection to AI, it is clear that we are no longer talking about just another technology, but about a subject with enormous human, social and ethical implications.
I read it not only from a spiritual perspective, but also from a business and societal one: what kind of progress are we building, what responsibilities do those of us developing technology have, and how can we ensure that the pursuit of efficiency never makes us lose sight of people?
One idea, in particular, stayed with me: AI is not only about technical capabilities—it is about judgement. And that is precisely why it matters who builds it, with what intention and for what purpose.
Business is built on…
…solving real problems for real companies, supported by the best team possible. Everything else—the technology, the business model and the brand—are simply tools.
In a highly technical industry like ours, credibility is not earned through great presentations alone. It is earned in the laboratory, by delivering measurable value from the very first week and by building long-term relationships.
When customers genuinely feel that their problem has become your problem, business naturally follows.
