Patina, a young fragrance-technology startup working on AI-assisted scent discovery, has raised $2 million from investors including Betaworks and True Ventures, bringing new venture funding to a niche field at the intersection of machine learning, chemistry and consumer-product design.

The company, led by Sean Raspet and Laura Sisson, is trying to build technology that can help identify and synthesize new scent molecules. Its approach combines molecular design, machine learning and perfumery, according to the company’s public materials and recent coverage of the financing. Patina describes its broader goal as understanding scent at the biological level by mapping how molecules activate receptors involved in smell, taste and sensation, then using that foundation to discover compounds that have not existed before.

The $2 million round is small by late-stage AI standards, but it is notable because it applies investor interest in foundation-model-style systems to a physical industry rather than a purely digital one. Over the past two years, venture capital has heavily favored companies building AI infrastructure, coding tools, enterprise agents and domain-specific models. Patina’s pitch belongs to a more specialized category: using computational systems to explore chemical space and produce candidates for real-world products.

The fragrance market depends on molecules that can be used in perfumes, cosmetics, household products, candles and other scented consumer goods. Those ingredients are often developed by specialized laboratories and fragrance companies, then incorporated by brands into finished products. TechCrunch, citing Patina’s view of the market, reported that traditional fragrance developers may examine large numbers of candidate molecules before only a small fraction reach perfumers’ working palettes.

Patina is positioning itself against that slow funnel. The startup says it is not simply trying to refine existing fragrance ingredients, but to build a system for generating what comes next. Its website says the company is “mapping how molecules activate the receptors that produce smell, taste, and sensation,” and then using that knowledge to “discover and synthesize compounds that have never existed before.”

That framing places Patina in a broader wave of startups and research teams applying AI to science and materials discovery. In drug development, protein design and battery chemistry, investors have backed companies that use machine learning to reduce experimental search costs and identify promising candidates earlier. Fragrance is a smaller and more sensory market, but it presents a related challenge: predicting how molecular structure translates into human perception.

Unlike a text model, a scent model must contend with biology, chemistry, formulation constraints and subjective human experience. A molecule that looks promising in software still has to be synthesized, tested for safety, evaluated for stability and accepted by perfumers, brands and regulators. That gives Patina a harder commercialization path than many software startups, but also a clearer reason to exist if its technology can materially improve the discovery process.

The company’s co-founders bring backgrounds that fit the hybrid nature of the problem. Patina’s public LinkedIn profile lists Raspet as co-founder and CEO and Sisson as co-founder and CTO. The company describes itself as operating in biotechnology research, with specialties including fragrance, flavor, aromachemicals, molecule design, cosmetics, foundation models, AI, machine learning and simulation. It lists Brooklyn, New York, as its headquarters and says it was founded in 2025.

Betaworks has also previously described Patina in its AI Camp materials as a company working on a model for scent and flavor. That accelerator-style description framed the opportunity around using deep learning and receptor modeling to create safer and more distinctive fragrance molecules, as well as potential substitutes for natural ingredients that can be costly or resource-intensive to produce.

The sustainability angle could become important if Patina’s technology can help create synthetic alternatives to ingredients traditionally derived from plants or other natural sources. Some natural fragrance inputs are expensive, supply-constrained or environmentally burdensome when scaled. A molecule that recreates a desired scent profile with lower production impact could appeal to brands looking to reduce exposure to volatile commodity inputs or build more consistent products.

Still, the startup faces a market in which incumbents have long-standing relationships with fragrance houses, beauty companies and consumer-goods manufacturers. Established players have deep libraries of ingredients, regulatory expertise and global formulation teams. Patina’s challenge will be to show that its model can produce molecules that are not only novel, but useful, safe, scalable and desirable to professional perfumers.

The raise also comes as AI investors are searching for defensible applications beyond general-purpose chatbots and productivity tools. A startup building proprietary datasets, domain expertise and physical-world outputs can offer a different kind of moat than a software interface layered on top of existing large language models. In Patina’s case, that moat would likely depend on the quality of its scent data, its receptor-modeling approach, its ability to synthesize candidates and its relationships with fragrance buyers.

For the technology industry, the financing is another reminder that AI’s commercial frontier is spreading into sectors that have historically moved through lab work, specialist knowledge and slow product-development cycles. Patina is not yet a large company, and the newly disclosed round does not by itself prove that AI can transform fragrance creation. But it gives the startup early backing to test whether computational models can make the search for new scent molecules faster, cheaper and more systematic.

If Patina succeeds, its most important products may never be consumer apps or branded perfumes. They may be ingredients that sit inside other companies’ products, invisible to buyers but central to how those products smell. That makes the company’s work less visible than most Silicon Valley AI launches, but potentially meaningful in a consumer economy where sensory differentiation remains a powerful commercial tool.