Biomatter Is Delivering a Custom Enzyme Like a Catalog Product

The Lithuanian startup’s bet is that full-stack design and delivery can turn bespoke protein engineering into an order-and-deliver business.

About Biomatter

Published

Biomatter’s commercial pitch is a simple one: a custom enzyme should be as easy to order as a catalog product. For the Vilnius-based startup, that means taking a customer’s chemical reaction, designing a novel protein to catalyze it using AI and physics models, and then shipping the physical enzyme. The company is not selling software or a design service. It is delivering a functional, tested molecule, including for reactions where no natural enzyme exists [VU.lt, November 2026].

This full-stack approach, which combines computational design with wet-lab validation, has landed the company more than 25 enterprise customers across biopharma, agriculture, and industrial biotechnology. Its enzymes are already in products from AstraZeneca, BASF, and Sanofi, and its work is incorporated into a Thermo Fisher Scientific reverse transcriptase used in mRNA research [VU.lt, November 2026] [Businesswire, Unknown]. In November 2026, Biomatter closed an $18 million Series A round led by Astanor to expand its commercial team and build out a dedicated design-and-testing facility in Vilnius [Vestbee, November 2026].

From Academic Breakthrough to Industrial Foundry

The company’s technical foundation was laid in a 2019 paper in Nature Machine Intelligence, where the founding team demonstrated the first fully functional enzymes created with generative AI [biomatter.ai, August 2024]. That research, conducted at Vilnius University’s Institute of Biotechnology, proved that AI models could design viable protein sequences far outside the known natural space. Biomatter was formally founded in 2018 to commercialize this capability, with co-founders Laurynas Karpus, Vykintas Jauniškis, and Irmantas Rokaitis leading the transition from academic lab to industrial foundry [Move the Needle, November 2026].

Their core insight was that the real bottleneck for industrial customers wasn’t just better design algorithms, but the entire workflow from specification to a shelf-stable enzyme. By owning both the computational platform and the laboratory testing, Biomatter aims to compress development cycles from years to months. The company reports it has delivered more than 50 custom enzymes to date [VU.lt, November 2026].

The Full-Stack Wedge in a Fragmented Market

The industrial enzyme market is historically fragmented. Large chemical companies often maintain internal R&D teams, while biotech startups might offer niche design tools or contract research services. Biomatter’s wedge is to own the entire value chain, positioning itself as a turnkey supplier.

  • Design control. The company’s proprietary platform, which includes tools like the LYNX Enzyme Optimizer, uses a combination of generative AI and physics-based simulations to explore a vast sequence space and predict stability and activity [biomatter.ai, Unknown].
  • Validation ownership. Every designed sequence is synthesized and tested in Biomatter’s own labs, closing the loop between digital design and physical performance. This integrated feedback is critical for iterating on difficult targets.
  • Commercial simplicity. The customer provides a target reaction; Biomatter handles the rest, delivering a characterized enzyme. This model appears in partnerships like the one with Neogen Corporation, announced in August 2025, to develop enzyme-based food safety products [Neogen Corporation, August 2025].

Traction Beyond the Lab Bench

Biomatter’s customer roster and partnerships signal early commercial validation in demanding, regulated industries. The inclusion of its enzyme design in Thermo Fisher Scientific’s Invitrogen SuperScript V Reverse Transcriptase is a particularly strong signal, as it demonstrates performance in a high-precision, commercialized life sciences product [TechFundingNews, November 2026].

The company has also joined the AstraZeneca BioVentureHub, a program designed to accelerate external innovation, and counts Kirin Holdings and Tereos among its clients [biomatter.ai, November 2026] [VU.lt, November 2026]. This geographic spread,with customers in Europe, the United States, and Japan,suggests the value proposition translates across major industrial regions.

2022 Seed | 2.1 | M USD
2026 Series A | 18 | M USD

The Technical Breakdown: Where the Friction Lives

From an infrastructure perspective, Biomatter’s model introduces specific scaling challenges. The computational workload for generative protein design is immense, but it’s the wet-lab throughput that becomes the ultimate governor. Each designed enzyme requires synthesis, expression in a host organism, purification, and functional characterization. This creates a physical queue. The new Vilnius facility funded by the Series A is essentially a capacity play to increase the number of design-test cycles it can run in parallel [TechFundingNews, November 2026].

The sober assessment is that the company’s asset-heavy model could face pressure at scale. While the integrated approach is the source of its moat today, it also means growth is tied to capital expenditure on lab space, equipment, and specialized personnel. The unit economics of designing and delivering a one-off enzyme for a novel reaction are not the same as scaling a software product. A misstep in forecasting demand or a delay in lab throughput could quickly erode margins. The bet is that the premium for a guaranteed, custom enzyme will justify the operational complexity and that process innovations will steadily drive down the cost per designed protein.

What to Watch in Vilnius

The next twelve months will test Biomatter’s capacity expansion. The primary milestone will be the operational launch of its new Vilnius design-and-testing facility and the subsequent ramp in project throughput. Commercially, the focus will be on deepening relationships with existing enterprise customers to move from one-off projects to recurring enzyme supply agreements, a key step in building predictable revenue.

The company’s partnership with Neogen, aimed at setting new industry standards for efficiency [Investing.com, Unknown], will be a live case study for its model in the food safety sector. Success there could open similar verticals in agriculture and environmental remediation. For a company betting it can turn bespoke bioengineering into a repeatable service, the proof will be in the consistent, profitable delivery of the next fifty enzymes.

Sources

  1. [VU.lt, November 2026] Vilnius University-born startup Biomatter raises 18 million in investment | https://www.vu.lt/en/all-news/vilnius-university-born-startup-biomatter-raises-18-million-in-investment-1
  2. [Vestbee, November 2026] Biomatter raises $18M Series A to accelerate AI-powered enzyme design | https://www.vestbee.com/insights/articles/biomatter-raises-18-m
  3. [Move the Needle, November 2026] Biomatter raises $18 million to speed up AI enzyme design | https://www.movetheneedle.news/start-ups/biomatter-raises--18-million-to-speed-up-ai-enzyme-design/
  4. [Businesswire, Unknown] Thermo Fisher Scientific press release industry page | https://www.businesswire.com/newsroom?industry=1778606
  5. [biomatter.ai, August 2024] Biomatter company page | https://biomatter.ai/
  6. [Neogen Corporation, August 2025] Biomatter Establishes Relationship with Neogen | https://biomatter.ai/resources/news/biomatter-establishes-relationship-with-neogen/
  7. [TechFundingNews, November 2026] Biomatter lands $18M from Astanor and UVC Partners to build a Vilnius enzyme lab | https://techfundingnews.com/biomatter-lands-18m-from-astanor-and-uvc-partners-to-build-a-vilnius-enzyme-lab/
  8. [biomatter.ai, November 2026] Biomatter joins AstraZeneca BioVentureHub | https://biomatter.ai/resources/news/biomatter-joins-astrazeneca-bioventurehub/
  9. [Investing.com, Unknown] Article on Neogen and Biomatter partnership | https://www.investing.com/

Read on Startuply.vc