Mafer AI, a Barcelona-based startup constructing an AI working system for R&D groups in formulation industries, has closed a €2 million pre-Seed funding spherical backed by Kfund, 4Founders Capital, Masia and Lavanda Ventures, the startup funding arm of the Puig household.
It has additionally secured backing from main enterprise angels, together with Adrián Mato (Andreessen Horowitz scout and GitHub Copilot government), Manuel Roca (founding father of Atrápalo and investor in Wikiloc), Dídac Lee, Fernando Castiñeiras (managing accomplice at Athos Capital), and Juanjo Mostazo, alongside traders from the cosmetics and perfume industries and professionals from MIT, Google, Samaipata, Bain, BCG, Shakers and Pleased Robotic.
Fernando Oliver Jané, CEO and co-founder of Mafer AI, mentioned, “Formulation industries have spent many years accumulating a silent asset, their technical R&D historical past, with out having any technique to exploit it at scale. The present era of AI modifications that equation utterly: immediately, utilizing the info of a single buyer, we will prepare proprietary fashions that radically speed up the tempo at which that firm brings formulation to market.
“We’re constructing a brand new class from Europe, throughout a window that may stay open for the subsequent three years, and that we consider won’t reopen once more. With €2 million, a unprecedented group and the assist of the Barcelona Supercomputing Centre behind us, we intend to construct it ourselves.”
Based in Barcelona in late 2025 by Fernando Oliver Jané (CEO) and Marc Montalbo Burges (CTO), Mafer AI is the corporate behind MaferOS, an AI-native working system for R&D groups throughout formulation industries, together with specialty chemical compounds, meals, drinks, cosmetics, private care, and fragrances and flavours.
In accordance with the corporate, it’s focusing on a silent however structural bottleneck affecting formulation-based industries, which encompasses a broad vary of sectors, together with flavours and components for meals and drinks, fragrances for cosmetics and private care, in addition to pharma and nutraceuticals.
It states that corporations working in these sectors have collected many years of technical know-how of their laboratories on how profitable merchandise are constructed, within the type of failed formulation, laboratory analyses, regulatory approvals and technical choices. But this asset stays fragmented throughout closed laboratory devices, spreadsheets, regulatory PDFs and, most significantly, the minds of senior specialists.
Mafer AI factors out that this leads to the innovation cycle working between 5 and ten occasions slower than the aggressive tempo of the market, with delayed product launches and shrinking margins. It additional notes that when senior specialists retire or depart, corporations can lose as much as two years of institutional information. In the meantime, technical groups proceed to function on decade-old software program infrastructures, whereas reformulating a product nonetheless requires manually recalculating regulatory compliance throughout dozens of jurisdictions.
MaferOS combines synthetic intelligence fashions with a proprietary structure tailored to every buyer, coaching proprietary fashions on every firm’s historic knowledge whereas maintaining all data protected and remoted.
The corporate defined that the platform is structured into specialised modules able to automating technical choices which might be at present dealt with manually or by means of fragmented workflows, from laboratory evaluation and knowledge structuring to regulatory compliance and formulation advice.
“What differentiates Mafer’s method is just not merely making use of AI to chemistry, however bringing to the sector the full-stack enterprise software program mannequin already reshaping different industries: proprietary fashions skilled per buyer, AI brokers orchestrated on prime of structured knowledge layers, and a group of Ahead Deployed Engineers embedded instantly inside shopper organisations to carry merchandise into manufacturing inside weeks,” the Spanish startup talked about within the press launch.
It asserts that it’s the identical working mannequin popularised by Palantir and not too long ago validated by corporations corresponding to Pleased Robotic and Fantastic in enterprise environments. Mafer claims to be the primary firm to carry this method to specialty chemical compounds and FMCG.
Its platform permits proprietary fashions skilled on every shopper’s technical historical past whereas defending proprietary knowledge; specialised AI brokers that orchestrate complicated technical choices; computerized structuring of fragmented R&D historical past, together with chromatograms, technical sheets, formulation and regulatory documentation; automated regulatory compliance throughout dozens of jurisdictions; and manufacturing deployment inside weeks, all by means of a long-term partnership mannequin designed to maintain shoppers on the technological frontier.
Simply six months after launch, Mafer AI stories that it has already secured near €300k in dedicated ARR and is focusing on €1 million ARR by the top of 2026. The corporate at present has three manufacturing shoppers within the perfume and flavours vertical, the place it has structured extra
than 50,000 technical paperwork. It additionally highlights sustaining an energetic enterprise pipeline throughout six trade verticals, with proof-of-concept initiatives underway in 4 of them. The startup has moreover gained three aggressive procurement processes in opposition to worldwide consulting corporations.
Mafer AI has additionally been chosen by the Barcelona Supercomputing Centre for its AI Manufacturing facility programme (batches 1 and a couple of), granting the startup entry to greater than 50,000 GPU hours on NVIDIA H100 infrastructure and the MareNostrum V supercomputer.
The corporate plans to make use of this capital to broaden the technical group by hiring further founding engineers and strengthening each the Ahead Deployed Engineering and analysis features. The group at present consists of eight individuals, together with the 2 co-founders, and plans to scale to 12 staff throughout the second quarter of 2026.
It is going to additionally deploy this capital to deepen relationships with current prospects and flip them into flagship public references whereas increasing new modules on prime of the deployed platform, and advance the corporate’s roadmap towards proprietary foundational fashions skilled on the Barcelona Supercomputing Centre.
