Caravel Bio Secures $1.2M NSF Phase II Grant to Scale the Data Engine Behind Its Full-Stack Protein Engineering Platform

The grant funds the next chapter for Caravel's full-stack platform—built to create value for all proteins—while scaling machine learning datasets and real-world pilots for existing customers. Michał Jastrzębski also joins as CTO to lead the company's AI-driven protein engineering work.

PORTLAND, Ore., Oct. 28, 2026 /PRNewswire/ -- Caravel Bio, the company creating new high-performance systems for discovering, engineering, and manufacturing biomolecules, has been awarded a $1,246,984 Small Business Innovation Research (SBIR) Phase II grant from the National Science Foundation (NSF). The award, secured by fewer than 5% of applicants over the last decade, will support development of a system for creating and scaling proteins with new-to-nature chemistries. The Caravel platform massively parallelizes biological experimentation, enabling the building and testing of millions of protein designs at once. That scale generates the data AI needs to engineer proteins nature never built, accelerating development from the lab to products that can strengthen the economy, reduce pollution, and improve human health. The company has also welcomed Michał Jastrzębski, formerly of Ginkgo Bioworks, Proxima, and Intel, as Chief Technology Officer to lead the company's work in training advanced AI models for biomolecular design.

Nearly every industrial enzyme, therapeutic protein, and biomaterial on the market today has been built from the same 20 canonical amino acids found in nature. Beyond those 20 exist hundreds of additional amino acids that could enable amazing new products—for example, the blockbuster drug Ozempic improves stability using one non-canonical amino acid—but, as a field, biotechnology has generally lacked reliable methods to discover proteins and manufacture products at scale using these new chemistries. And while AI is generating promising protein designs faster than ever, labs can test only a fraction of them. Compounding the problem, protein engineers don't have effective feedback loops for turning laboratory experiments into scalable products without enormous expense and risk.

Caravel is building systems to close these gaps simultaneously through its novel design-build-test-learn cycle.

"I was really exhausted by the inability of the work I and my peers were doing, all with the best intentions, to actually improve the world without spending hundreds of millions of dollars to get a product across the finish line," says Trevor Nicks, PhD., Founder & CEO of Caravel. "So when we set out to build Caravel, we started completely from scratch, re-designing the entire process of biomolecular engineering to lower costs while increasing speed, volume, and the real-world relevance of generated data."

Caravel's platform combines a suite of new lab techniques that unlock the broader amino acid palette for both discovery and manufacturing of new proteins. This data-rich workflow incorporates machine learning that converts each round of experiments into sharper predictions for the next. Notably, the data generated not only identifies what should be made, but also lays the foundation for manufacturing it at scale.

"When an experiment costs a fraction of a cent instead of tens of dollars, you stop asking which handful of ideas might work and instead start testing all of them," says Emily Bowman, PhD., CSO of Caravel. "The systems we have built unlock massive data sets with very low costs. AlphaFold's success in protein design illustrates the potential of large datasets. But importantly, those datasets were made possible by billions of dollars in publicly funded research. We are thankful that the NSF recognizes the value in decreasing those costs and is supporting our work in the massive parallelization of experimentation."

One Year In: Partnerships, Funding, and Performance Gains
Since completing its Phase I grant in 2024, Caravel has translated its lab-stage platform into industrial partnerships, funding, and measurable performance gains. The company:

• Formally launched in October 2025 alongside a $7.8 million award from the NSF Directorate for Technology, Innovation and Partnerships
• Joined the Shell GameChanger program to create new ultra-stable enzymes for low-cost carbon capture and surpassed a key technical milestone within six months by increasing enzyme stability under industrially relevant conditions 12-fold
• Built a network of scientific and strategic partners, including Avery Bio, Caltech, Rutgers University, Oregon State University, Southern Connecticut State University, Rubi Laboratories, and the vHive Animal Health Incubator

Under Phase II, Caravel aims to extend its platform to additional enzyme classes for existing customers, and eventually scale those products for pilot deployment to innovate new, real-world protein products across DNA writing, carbon capture, rare earth separations, and more.

Deepening Leadership: Caravel's New CTO
The company also announced the addition of Michał Jastrzębski as Chief Technology Officer. As Caravel builds the infrastructure poised to define the next era of protein design, Jastrzębski will lead Caravel's artificial intelligence initiatives while helping advance its product roadmap and market strategy. Jastrzębski began his extensive career in technology at Intel Corporation and has since held roles at GitHub, bytewax, and Promixa. More recently, Jastrzębski worked at Ginkgo Bioworks, where he co-authored a collaboration with OpenAI exploring the use of a GPT-5-driven autonomous laboratory to optimize cell-free protein synthesis.

"I've seen firsthand how much AI in biology is still held back by a lack of the right data, not a lack of ambition," said Jastrzębski. "Caravel is solving that problem at the source. Joining as Chief Technology Officer means helping build the system that finally gives protein design the scale of data it needs."

ABOUT CARAVEL BIO
Caravel Bio has built a new data engine for biomolecular engineering, dramatically increasing the scale and speed of generating data. The platform tests millions of protein designs in parallel at a fraction of a cent per experiment, versus tens of dollars with conventional methods, providing the data AI-driven biology has been missing. The company's Artificial Evolution™ technology combines cell-free protein synthesis, microfluidic droplets, and bacterial spore display to supercharge biomolecular engineering, so labs can generate the data AI needs to design proteins nature never built. Caravel was selected to lead a $7.8M NSF Technology, Innovation and Partnerships award and a $1.2M NSF SBIR Phase II grant. Caravel is on a mission to expand the role of biology as a transformative force for societal good. Learn more at www.caravel.bio.

About the National Science Foundation's Small Business Programs
America's Seed Fund powered by NSF awards $200 million annually to startups and small businesses, transforming scientific discovery into products and services with commercial and societal impact. Startups working across almost all areas of science and technology can receive up to $2 million to support research and development (R&D), helping de-risk technology for commercial success. America's Seed Fund is congressionally mandated through the Small Business Innovation Research (SBIR) program. The NSF is an independent federal agency with a budget of about $8.5 billion that supports fundamental research and education across all fields of science and engineering.

MEDIA CONTACT
Tess Pawlisch
608-333-9788
[email protected]

SOURCE Caravel Bio