Patina is a fragrance tech startup building new scent molecules with molecular design, machine learning, and sensory science, and it has now raised $2 million from investors including Betaworks and True Ventures. The pitch is simple, even if the science isn’t: a small set of specialist labs still controls a lot of how fragrance ingredients get invented, while natural materials keep getting harder and pricier to source. Sean Raspet and Laura Sisson started Patina after meeting in New York in 2024 and formally launching the company in 2025. Now they’re trying to turn smell into something closer to a programmable system than a black-box craft.
What is Patina and how does the fragrance tech startup work?
Patina’s core product is a foundation model called Sense1. It predicts how a molecule will activate the human olfactory receptor system, then maps those receptor patterns to perceived smell. In plain English: instead of asking whether a molecule sounds vaguely “floral” or “woody,” Patina tries to model what’s happening one layer deeper — at the receptor level — across roughly 400 human olfactory receptors.
That changes the workflow for a customer. A fragrance house or brand can start with a target scent profile or a natural material it wants to reproduce. Patina can screen candidate molecules against receptor activity patterns tied to that smell. From there, the company can generate custom ingredients or propose alternatives to scarce natural inputs. It can also build “never-before-smelled” molecules that still hit a specific sensory brief. Patina says receptor-level modeling lets it replicate materials like rose oil without relying on plant extraction.
Sense1 isn’t just a pitch deck idea. Patina’s white paper says the model predicts receptor activation across the full human receptor set and outperforms older docking-based methods by a wide margin. The company reports an AUROC of 0.854, a 39% improvement over optimized structure-based baselines, and says its top-1 screening hit rate reaches 54% versus 3.3% for the docking pipeline it tested against. Those are the numbers investors care about. They suggest the model may cut down failed screening work in the lab.
The manual work Patina wants to remove is the fuzzy part of fragrance development. A lot of scent description still depends on human language, taste, culture, and regional shorthand. Sense1 is built around the idea that biology is a more stable reference layer than words. Patina even frames the model as a step toward “the first universal code of smell and taste,” an ambitious claim, but one attached to a real technical thesis.
How did Patina start, and who built this fragrance tech startup?
The founding story
This wasn’t a typical startup origin story. Raspet and Sisson met at a scent art gallery in New York in 2024 — Raspet was showing new molecules, and Sisson was building olfactory learning models. They started collaborating on research, then turned that work into Patina in 2025. Raspet’s line about the moment is still the cleanest summary: it felt like “the timing was right to finally build the tools to understand scent at the biological level.”
Why the founders make sense for this
Raspet brings a weird but very relevant mix of art, perfumery, and formulation work. Long before Patina, he was already known for projects built around flavor and fragrance molecules. He also worked as a flavorist at Soylent and co-founded Nonfood, an algae-based food company. That matters. Patina isn’t being built by someone who discovered smell as an AI prompt last month. Raspet has been working around sensory chemistry for years.
Sisson’s path is different, but it lines up. She came from food and software engineering, and her research record backs up the smell-modeling part. She has published work on odor descriptor understanding and predicting the smell of aroma-chemical pairs. More recently, she benchmarked how language models reason about olfaction. That gives Patina something a lot of startups don’t have: a founding team split between molecule intuition and computational modeling.
Early signals, product status, and the round itself
Patina is still early. But it’s not just theory. The company is already in talks with top fragrance houses and fashion brands about custom scent work, and the new capital has helped it move from Raspet’s backyard into an office in Bushwick, Brooklyn, where it now has a small group of chemists. Sense1 is also being made available to select research and industry partners. That suggests Patina is treating the model as infrastructure, not just internal tooling.
On the financing side, Patina raised $2 million, with investors including Betaworks and True Ventures. Raspet says the money is going toward launching new molecules and building new partnerships. It’s also funding receptor activation data collection through startup and academic lab collaborations. That last part is a big deal. For a company like this, data isn’t a nice-to-have. It’s the moat.
How Patina compares with Osmo and the old guard
Patina isn’t alone. Osmo is the best-known startup in digital olfaction right now, and it raised $70 million in Series B funding in early 2026 to expand fragrance ingredient design and manufacturing. Osmo pairs AI formulation with in-house production. It launched its Generation ingredient arm in March 2025 and is pushing on lower minimum orders and faster development cycles. That’s a much more industrial posture than Patina’s current stage.
Then there are the incumbents. Givaudan, Symrise, DSM-Firmenich, and IFF already use AI in fragrance R&D and still dominate the ingredient pipeline through scale, data, manufacturing, and customer relationships. Patina’s opening is narrower. It’s betting that receptor-level biology can produce more exact replications and more novel ingredients. It could also produce more patentable building blocks in a market where formulas themselves usually aren’t patent protected. That’s smart. But it’s also hard. Great models are one thing. Shipping molecules that perfumers actually keep buying is another.
Why does Patina’s $2M round matter?
This round looks less like a splashy valuation exercise and more like a conversion point.
Patina was already doing the science. The fresh money gives it the basics needed to turn that science into a repeatable business — lab space, chemists, partner programs, and more receptor data. For a company building a foundation model around smell, that’s the stack that matters most. Not Super Bowl ads. Not massive headcount. Data, experiments, and molecules.
There’s also a real customer implication here. Sisson says buyers increasingly want “newer, safer and more expressive perfumes,” and Patina is trying to answer that with custom ingredients that can be developed in weeks instead of years. If it works, smaller brands could get access to ingredient creation that used to be locked inside giant suppliers and long development cycles.
Investors aren’t just backing a perfume label. They’re backing the idea that Patina could become an underlying intelligence layer for scent design. That’s a much bigger swing. If the company can own a better map between molecules, receptors, and perception, it could matter in fragrance, flavor, cosmetics, and maybe even skin-related applications later on.
How big is the market for AI fragrance design?
The market is already large enough that even a niche infrastructure player can matter. Grand View Research estimates the global flavors and fragrances market at $33.58 billion in 2025 and projects it will reach $57.52 billion by 2033, growing at a 7.1% CAGR. In the U.S. alone, the market was estimated at $5.80 billion in 2023 and is forecast to hit $7.43 billion by 2030.
Demand is shifting in ways that fit Patina’s timing. Grand View says the natural segment held 51.6% of global revenue in 2024, while natural products made up 76.3% of the U.S. market mix in 2023. Consumers want cleaner and more premium ingredients. Brands want dependable supply. Those two things don’t always coexist when you’re depending on crops, weather, extraction yields, and commodity swings.
That’s why AI keeps showing up in scent R&D. Better protein modeling and better screening are starting to make fragrance design feel less mystical. Structure-to-perception prediction is getting better too. The entire category is inching in that direction, whether through startups like Osmo or through the big suppliers upgrading their internal tooling. Patina’s bet is that smell has been waiting for its equivalent of a color code.
Can this fragrance tech startup become a Pantone for scent?
That’s the right way to think about Patina.
Not as another perfume brand. Not even really as a software company in the usual sense. More like an attempt to build a reference system for smell that can feed ingredient discovery and custom formulation. It could also support synthetic replacements for fragile natural materials. Raspet calls it a “Pantone for scent,” and that’s the part worth watching.
The fragrance tech startup now has the first outside capital, a working scientific thesis, and early commercial conversations. What it still needs is proof that receptor-level modeling can consistently produce ingredients people want to buy at commercial scale. If that happens, Patina could matter far beyond perfume. If it doesn’t, it’ll still be one more reminder that smell is a brutal category to industrialize.
Read how Convective Capital raised an $85M fund to back resilience tech startups building tools for wildfire prevention, disaster response, infrastructure hardening, and climate risk management.
FAQ
– What funding did Patina raise?
Patina raised $2 million in 2026 from investors including Betaworks and True Ventures. The round is meant to support new molecule launches, research partnerships, and the collection of receptor activation data that can improve its underlying model.
– How does Patina’s Sense1 platform work?
Sense1 models how molecules activate human olfactory receptors and then links those activation patterns to perceived scent qualities. Patina says that lets it screen candidates more effectively and replicate rare natural ingredients like rose oil at the biological level. It can also design new scent molecules with more precision than descriptor-based methods.
– Who founded Patina?
Patina was founded by Sean Raspet and Laura Sisson after they met in New York in 2024 and launched the company in 2025. Raspet came in with deep flavor-and-fragrance experience from art, Soylent, and Nonfood, while Sisson had already been working on machine learning research related to odor language and aroma prediction.
– Is Patina in the perfume market or the fragrance ingredients market?
It’s closer to the fragrance ingredients market than to consumer perfume retail. Patina is building ingredients and modeling tools that could be used by fragrance houses, fashion brands, cosmetics companies, and flavor developers inside a broader global flavors-and-fragrances market that Grand View sizes at $33.58 billion for 2025.
