One To Watch

Profluent

Company Overview

A Berkeley-based AI biotechnology company applying large-scale foundation models to protein design, with a strategic focus on engineering custom site-specific recombinases for therapeutic gene editing applications. Profluent sits at the intersection of generative AI and molecular biology, using language model architectures trained on protein sequence data to design novel proteins beyond what nature has produced. The company's core thesis is that AI-designed proteins can unlock therapeutic categories that conventional protein engineering and CRISPR-based approaches cannot easily address.

Headquarters and Global Presence

Profluent is headquartered in Berkeley, California, operating within the dense biotech and academic ecosystem surrounding the San Francisco Bay Area. The company's proximity to UC Berkeley and related research institutions underpins its computational and scientific talent pipeline.

Founding and History

Profluent was founded with the explicit goal of making protein design a programmable, AI-driven discipline rather than an empirical one. The company gained significant public attention in 2024 when it published OpenCRISPR-1, an open-source AI-designed gene editor — one of the first publicly released AI-generated proteins demonstrated to function in human cells. This milestone positioned Profluent as a credible scientific actor in the emerging field of AI-generated biologics, attracting both academic interest and commercial partnership discussions.

Therapy Areas and Focus

Profluent's therapeutic ambitions center on diseases with severe unmet needs where precise genomic intervention offers potential curative benefit, including rare genetic disorders and select oncology applications. Site-specific recombinases are particularly attractive for conditions requiring large-payload DNA insertion or complex chromosomal rearrangements that exceed the editing capacity of nuclease-based tools. The company's collaboration with Eli Lilly signals a prioritization of disease areas where Lilly holds deep clinical and commercial expertise.

Technology Platforms and Modalities

Profluent's foundational platform is a suite of large-scale protein language models trained on vast databases of evolutionary and synthetic protein sequences, capable of generating novel protein sequences with specified structural and functional properties. The company applies this platform specifically to recombinases — enzymes that catalyze site-specific DNA recombination — designing custom variants that recognize predetermined genomic sequences and execute precise insertions, deletions, or inversions. Unlike CRISPR nucleases, site-specific recombinases do not require guide RNAs and leave no double-strand breaks, potentially offering a cleaner safety profile. The approach allows therapeutic proteins to be designed computationally and iteratively optimized at AI speed rather than through slow directed-evolution campaigns.

Key Pipeline and Programs

Profluent does not currently disclose named clinical-stage drug candidates, reflecting its position as a platform-stage company translating AI-designed proteins toward therapeutic development in partnership with larger organizations.

The lead collaborative program, announced in April 2026, involves a multi-program strategic research collaboration with Eli Lilly focused on developing and commercializing custom site-specific recombinases for diseases with severe unmet needs. Profluent's AI platform will generate recombinase candidates tailored to Lilly-specified genomic targets, with Lilly providing disease biology, development, and commercialization capabilities. Financial terms of the collaboration have not been disclosed publicly, but the multi-program scope indicates more than a single-indication proof-of-concept arrangement.

The OpenCRISPR-1 program, released as an open-source resource in 2024, remains the most publicly documented example of a functional AI-designed gene editor produced by the platform. While not a proprietary therapeutic asset, it serves as scientific validation of the company's ability to generate novel editors that function in human cells.

Recent Developments

The most significant recent milestone is Profluent's multi-program strategic research collaboration with Eli Lilly, announced on April 28, 2026, targeting custom site-specific recombinases for severe unmet-need indications. This deal represents the company's most prominent commercial validation to date, pairing its AI protein design platform with one of the world's largest pharmaceutical companies. The partnership structure — described as strategic and multi-program — suggests a long-term research alliance rather than a one-time licensing transaction.

Key Personnel

Ali Madani serves as Chief Executive Officer and co-founded Profluent after leading protein generation research at Salesforce Research, where he was a principal architect of early protein language model work that informed the company's platform. The broader leadership team draws from computational biology, machine learning, and synthetic biology, though additional executive appointments have not been publicly disclosed in available sources.

Strategic Partnerships

Profluent's collaboration with Eli Lilly, announced April 2026, is the company's flagship commercial partnership, covering multiple programs aimed at developing AI-designed site-specific recombinases for therapeutic use. The deal combines Profluent's generative protein design capabilities with Lilly's disease expertise, clinical infrastructure, and global commercialization reach. Prior to the Lilly deal, Profluent had demonstrated platform openness through the public release of OpenCRISPR-1, building scientific credibility that facilitated engagement with major pharma partners.

FAQ Section

Profluent's platform generates novel proteins computationally, but translating AI-designed recombinases into approved medicines requires deep disease biology expertise, clinical development infrastructure, and regulatory capabilities that a platform-stage company does not yet possess at scale. Partnering with Lilly allows Profluent to apply its AI design engine across multiple disease programs simultaneously while Lilly provides the therapeutic development and commercialization weight. The multi-program structure also provides Profluent with validation, resources, and real-world feedback loops to further train and refine its models.

Site-specific recombinases catalyze precise DNA recombination at defined genomic sequences without creating double-strand breaks, which are the primary source of off-target mutations and genotoxic risk associated with CRISPR-Cas nucleases. They can also facilitate large-payload DNA insertions — potentially entire therapeutic genes — more efficiently than nuclease-dependent homology-directed repair. The challenge has historically been that natural recombinases only recognize their native DNA sequences; Profluent's AI platform addresses this by designing custom recombinases programmed to recognize therapeutically relevant target sites.

Most AI drug discovery platforms focus on small molecules or apply AI to optimize naturally occurring proteins through directed evolution surrogates. Profluent's models are generative — they produce novel protein sequences not found in nature, analogous to how large language models generate novel text. The OpenCRISPR-1 release in 2024 was a deliberate public demonstration that AI-generated proteins can be functional in human cells, a credibility threshold that many competing platforms had not cleared with publicly verifiable data.

OpenCRISPR-1 is an AI-designed gene editor generated entirely by Profluent's protein foundation models and released as an open-source resource in 2024. It was among the first AI-created proteins demonstrated to achieve gene editing activity in human cells, providing independent scientific validation that the company's generative models can produce functional biological tools. By releasing it openly, Profluent also signaled a strategy of building scientific credibility and community engagement rather than locking all IP behind proprietary barriers.

The Lilly collaboration explicitly targets diseases with severe unmet needs, a phrase that encompasses rare genetic diseases and select oncology indications where site-specific genomic correction could deliver transformative benefit. Recombinase-based editors are particularly well suited to monogenic disorders requiring stable gene insertion, making inherited metabolic diseases and hematologic conditions natural targets. The multi-program collaboration structure with Lilly suggests the pipeline spans more than one disease area from the outset.

Profluent is a preclinical platform-stage company; no candidates have entered clinical trials as of early 2026. The Lilly collaboration is framed as a research collaboration focused on developing custom recombinases, indicating that lead candidate identification, in vivo validation, and IND-enabling studies remain ahead. Given typical timelines for novel gene editing modalities, initial IND filings from the Lilly programs would likely fall in the 2027–2029 window at the earliest, though platform breakthroughs or accelerated development pathways could shorten that horizon.

Key watchpoints for Profluent include:

Progress disclosures from the Lilly collaboration, including nomination of specific disease targets or recombinase candidates advancing toward IND-enabling studies.
Additional partnership announcements that would validate the platform's breadth across multiple disease areas and pharma partners.
Publication of in vivo efficacy and safety data for AI-designed recombinases in preclinical disease models, which would be a major scientific milestone.
Competitive pressure from other AI protein design companies and established gene editing players pursuing recombinase and transposase-based editing systems.
Financing milestones, as preclinical platform companies require sustained capital ahead of clinical proof-of-concept.

Want to Update your Company's Profile?