Profluent, a fast-growing player in protein engineering, has secured 106 million dollars in new financing co-led by Altimeter Capital and Bezos Expeditions. The round brings its total funding to 150 million dollars, giving the Emeryville, California-based startup fresh capital to accelerate its push to make biology “programmable” across therapeutics, diagnostics, agriculture and biomanufacturing.
Founded by machine-learning scientist Ali Madani, Profluent is one of the few companies applying large-scale language models directly to the design of new biological molecules. The company has already notched a series of scientific firsts: the demonstration that LLMs can generate functional proteins (Nature Biotechnology, 2023), the creation of the first AI-designed CRISPR system built entirely from scratch (Nature, 2025), and the establishment that scaling laws apply to protein design (NeurIPS spotlight, 2025).
At the core of its platform is the Protein Atlas, a collection of more than 115 billion unique proteins, which the company describes as the largest protein data resource in existence.
Profluent’s early breakthrough came through OpenCRISPR-1, its open-source gene editor trained on specialized protein language models. The tool is now used by thousands of commercial labs, academic groups and major pharmaceutical companies. That traction has allowed the company to expand its commercial scope into antibodies, antigens, enzymes, and other biologic formats.
“With this funding, we are accelerating the path from lab breakthroughs to real-world applications, creating bespoke proteins to solve challenges in human health and the environment,” Madani said. He added that the capital allows the company to broaden market reach while deepening its R&D pipeline.
Profluent has already signed commercial relationships with companies including Revvity, Corteva Agrisciences, The Rett Syndrome Research Trust, and Integrated DNA Technologies. Its genome-editing programs helped de-risk its platform as it positions itself as a partner for protein design across industries.
Investors See a Foundational Shift Toward Programmable Biology
Altimeter Capital partner Jamin Ball, who co-led the round alongside Jeff Bezos’s investment arm, said the company’s fusion of AI scale, biological insight and therapeutic development expertise positions it at the foundation of a new category. “Programmable biology opens the largest design space in nature and reaches into multi-trillion-dollar markets, and Profluent is only at the beginning of exploring it,” Ball said.
Madani began exploring the idea of large language models for biology in 2020 while at Salesforce, where he worked on ProGen, a generative AI protein-design project. “The same architecture used for English, you can use for biological languages like proteins,” he told Forbes. He left the company in 2022 and teamed up with University of Washington researcher Alexander Meeske to build Profluent around that vision.
The company’s models allow scientists to describe the properties they want in a protein—such as stability or manufacturability in natural language, and receive a DNA sequence designed to produce it. Madani believes that custom protein design, rather than discovery of natural proteins, is the more scalable path to breakthrough therapeutics and agricultural innovation.
Racing Toward Larger Models and First-in-Human Therapies
Profluent is building on its claim that scaling laws, the principle that model performance improves predictably with more data and compute apply to protein design. “One of the reasons Jeff was interested is that we have discovered scaling laws,” Madani said. The company recently introduced a new foundation model, Profluent E-1, which adds evolutionary context to protein design.
Profluent now counts partners such as Revvity, agricultural giant Corteva, and the venture-backed therapeutics company Ensoma. It says its valuation is approaching one billion dollars following the latest round.
The startup sits among a growing group of companies using generative AI to design biological molecules, including Isomorphic Labs, Xaira Therapeutics, and others focused on AI-first drug development. What differentiates Profluent is its explicit goal of creating new proteins from scratch, rather than searching for promising natural candidates.
Madani describes Profluent as being in the early “GPT-1 or GPT-2 era” of AI-enabled biology. Yet the company is already seeing real-world adoption. Thousands of users now rely on OpenCRISPR-1, and Profluent is working with partners in therapeutics, diagnostics, biomanufacturing, and agriculture.
In explaining the company’s purpose, Madani often draws a contrast between biological discovery and engineered systems. While people easily grasp rocket landings or electronics, he argues that proteins, “molecular machines” responsible for everything from vision to immune response are not fundamentally different in complexity. Historically, he notes, breakthroughs have depended on chance: Alexander Fleming’s accidental discovery of penicillin or the identification of CRISPR sequences in a Danisco yogurt facility.
What Profluent seeks is a shift away from “caveman-like” trial-and-error toward direct, AI-driven design. “If we can have a machine that can truly make biology programmable, we will have a conveyer belt of blockbuster solutions,” Madani said. But he maintains that for now the field is still in early development, even as commercial traction grows.
