mafer AI has closed a €2 million round to build software that turns scattered lab records into usable intelligence. The Barcelona startup announced the raise on 29 May 2026, and it lands at a moment when formulation R&D teams are looking for better ways to use the data they already own. If you build products in chemicals, cosmetics, food, or fragrances, this one is worth a read.
What mafer Builds:
mafer AI develops MaferOS, an AI operating system made for research teams in formulation-driven industries. The idea is simple. Decades of formulas, lab results, regulatory files, and expert know-how usually sit in spreadsheets, PDFs, and people’s heads. mafer captures that information, structures it, and makes it ready for machine learning models to work with.
The platform keeps each customer’s data separate and private. According to the company, every client’s data trains only that client’s own models, and it is never shared or reused for anyone else. That matters because formulation records are some of the most valuable intellectual property a chemicals or pharma company holds.
Three Core Modules:
MaferOS ships with specialized tools rather than one generic assistant. The chromatography module focuses on structuring analytical data. The formulation module uses generative models to support formula recommendations. The regulation module handles compliance work through agentic workflows.
Together, these modules support lab analysis, data management, formula suggestions, and technical decisions. mafer says more modules across chemical processes are on the way, which fits the operating system framework. The pitch is that your technical history becomes the system your team runs on, so chemical knowledge compounds over time instead of leaving when an expert retires.
Who Backed the Round:
The 2 million euro round drew a strong investor group. Kfund, 4Founders Capital, Masia, and Lavanda Ventures led the institutional side. Lavanda Ventures is the startup investment arm of the Puig family, the group behind fragrance and beauty brands like Paco Rabanne and Carolina Herrera, which makes it a fitting backer for a formulation-focused platform. Angels and operators from MIT, Google, Bain, BCG, Samaipata, Shakers, and Happy Robot joined as well.
mafer AI was founded in late 2025 by Fernando Oliver Jané and Marc Montalbo Burges. The founding team mixes mathematics, AI research, and applied chemistry, with backgrounds spanning MIT research, venture capital, and analytical chemistry labs. The capital will go toward developing the platform, growing the team, and building partnerships across formulation-heavy industries.
Why the Timing Works:
Formulation industries face two pressures at once. They need to launch products faster, and they need to manage growing regulatory requirements. A specialty chemicals AI platform that structures historical knowledge and embeds compliance into the workflow speaks directly to both needs.
There is also a practical reason this approach makes sense now. Companies can train proprietary models on their own historical data, which unlocks patterns that were hard to reach before. Founder Fernando Oliver Jané framed it clearly, saying formulation industries have spent decades building a valuable asset in their technical R&D history and that AI now lets them turn that history into proprietary intelligence.
Where mafer Fits:
Barcelona has grown into one of the most active AI hubs in Southern Europe, with steady venture funding flowing into local startups. mafer AI plans to build a global business from that base, and the founders say they intend to lead this category from Europe.
The startup also gets a compute edge close to home. mafer AI was selected for the BSC AI Factory programme at the Barcelona Supercomputing Center, across its first and second batches. That selection grants access to more than 50,000 GPU hours on NVIDIA H100 infrastructure and the MareNostrum V supercomputer, which gives the team serious resources to train customer-specific models.
For founders and operators reading this, the interesting signal is the wedge. Instead of selling a broad AI tool, mafer enters through chromatography, formulation, and regulation, three areas where formulation R&D teams feel real friction every day. That focus on a specific industry and a specific data problem is a clear product strategy.
What mafer AI Means for R&D Teams:
mafer AI is a clean example of vertical AI applied to a data-rich industry. The product turns fragmented chemical data into structured, model-ready intelligence, the funding gives it room to expand, and the focus on formulation R&D keeps the value concrete. For teams in chemicals, cosmetics, food, fragrances, and nutraceuticals, MaferOS offers a path to use the knowledge they already have. The early backing suggests investors see the same opportunity.
