Home Technology Team FAQ Get in touch

Pushing Cell
& Gene Therapy
over the brink.

Programmable integrases, a nuclease-free platform for precise, large-scale genomic integration.
No cuts, no limits.

The missing piece

Brink is engineering the missing piece in Cell & Gene Therapy.

For many diseases, we know what genetic instructions to add to patients' cells to cure them at the root, but lack the tools to do it reliably. Precise insertion of therapeutic genes at desired sites in the human genome would enable more broadly applicable genetic cures to multiple currently incurable diseases, but has eluded the field for decades.

Why integrases

Nature has already evolved a system for precise genomic insertion.

Nature has been solving this problem for 4 billion years.

Integrases are molecular machines used by viruses to insert themselves into their host's genome with no cuts, no collateral damage, and no reliance on the cell's own repair machinery.

The catch: natural integrases have evolved to recognize unique sequences in bacterial genomes. Retargeting them to new sites in the human genome is an intricate protein design challenge. We solve it using the same force that created them, evolution.

A hypothetical fully reprogrammable recombinase would be, in some respects, the ultimate genome editing agent.

Prof. David Liu, Harvard University, inventor, CRISPR Base Editing & Prime Editing
Select a property to compare tools
Platform

The power of evolution.
Squared.

We harness the breadth and depth of natural evolution, the focus of directed evolution, and the predictive power of generative AI to engineer therapeutic-grade integrases for any genomic target.

Explore the platform
FAQ

Common questions, straight answers.

Unlike CRISPR-Cas9, Brink's integrases do not induce double-strand breaks (DSBs). DSBs are the primary source of genotoxicity with nuclease-based editors. Our enzymes insert large payloads (>5 kb) at pre-defined genomic addresses with high precision, no off-target indels, and no reliance on error-prone DNA repair pathways.
Most integrase platforms rely either on directed evolution alone or AI alone. Brink is the only platform combining metagenomics, directed evolution at billion-variant scale, and a generative AI model trained on functional data, creating a closed loop that compounds with every cycle.
Large Serine Integrases can integrate payloads exceeding 10kb, well beyond the capacity of CRISPR-based tools and sufficient for full-length gene replacement in most therapeutic contexts.
Brink-Locate, our computational engine, systematically maps naturally occurring integrase-compatible sequences within any genomic region of interest, giving us multiple high-quality insertion site candidates for every therapeutic application.
Our priority areas are hematological cancers (CAR-T), monogenic rare diseases requiring large gene replacement, and in vivo gene therapy where safety margins with nuclease-based tools are insufficient.
Large Serine Integrases are compact enough to be packaged in standard vectors, both viral and non-viral, without the payload constraints that make bulkier editors difficult to deliver in vivo. Their nuclease-free mechanism also removes a key safety barrier for systemic delivery.
We are a seed-stage company with validated platforms and proprietary integrase assets in active development. Our lead asset BTx-1 is progressing toward preclinical proof of concept in human cells.
Many protein design models are trained on sequence alone. Brink-AI learns from both sequence and labeled functional data. It is trained on billions of integrase variants whose activity has been screened directly in the lab. This makes it capable of both reprogramming an integrase to a new genomic address and optimizing it toward therapeutic-grade performance.
Brink-Evo screens up to one billion integrase variants per day using a proprietary cell-free droplet platform. This scale is what generates the functional training data that makes Brink-AI possible, and what allows us to find rare, high-performance variants that would be undiscoverable at lower throughput.
Yes. We are open to platform licensing, co-development agreements, and research collaborations with pharma and biotech partners working on cell and gene therapy programs. Reach out via the contact form.
Press

In the press.

Backed by:
Contact

Let's talk.

Investor? We're building the definitive insertion platform for cell & gene therapy, and we're raising.

Pharma or biotech partner? Your sequence, our integrase. Open to licensing and co-development agreements.

Scientist? We're a team of 13 at the frontier of directed evolution, metagenomics, and protein AI. Come build with us.

Four engines.
One closed loop.

The only platform that identifies optimal genomic insertion sites and engineers the integrases to reach them.

Our platform

Built to compound.

Each engine feeds the next, expanding our starting points, generating training data at scale, and turning that data into instant design capability.

Select a card to explore the technology in depth
Brink-Locate
Our computational engine maps integrase-compatible sites within any genomic region of interest.
100%
hit rate across all genomic regions tested
Brink-Mining
Explores macrodiversity, scanning metagenomes to discover millions of natural integrases across deep evolutionary branches, with >60% wet-lab validation.
>1M
integrases discovered & counting
Brink-Evo
Explores microdiversity, generating & screening up to 1B engineered variants per day around promising natural starting points, using a fully cell-free droplet platform.
1B / day
engineered variants screened
Brink-AI
Learns from both Nature and the lab, trained on orders of magnitude more functional data than any existing model.
#1
largest functional integrase training dataset in existence
Applications

Therapeutic focus areas.

The genetic instructions for curing many diseases are already known, proven in approved therapies. What's been missing is a way to add them precisely and safely into the right places in the genome, at the scale patients need. That's what we build.

01

CAR-T · Blood cancers

Precise, large-payload integration of CAR constructs into defined safe-harbor loci in T cells. Enables safer manufacturing, consistent expression, and next-generation multi-antigen designs.

02

Rare genetic diseases

Durable correction of monogenic diseases through site-specific integration of full-length functional gene copies, without genotoxic double-strand breaks.

03

In vivo delivery · Systemic treatments

Certain tissues and disease contexts demand delivery directly in the body, where precision requirements exceed what CRISPR can reliably deliver. Brink integrases are purpose-built for this: small enough to package efficiently, and nuclease-free by design.

Partner with us

Your sequence, our integrase. We engineer insertion enzymes tuned to any genomic address, open to licensing & collaborations.

Get in touch →
Brink team

Built by scientists.
Driven by patients.

About

Building the insertion platform
that gene therapy has been waiting for.

Brink Therapeutics is a Paris-based TechBio company founded in 2024. We combine directed evolution, metagenomics, and generative AI to engineer programmable integrases, a new class of genomic insertion tools that integrate precisely, safely, and at payload sizes that current tools cannot match. Backed by a €6M seed from Kurma Partners, Breega, Kima Ventures, PlugAndPlay, and BPI France.

Co-founders

Same team, new problem.

Jonathan, Harry, and Andrew first came together around one of biotech's most urgent challenges: engineering the molecular machinery behind Detect's Covid-19 diagnostic tests, shipped from concept to clinical scale in under a year. When that chapter closed, they kept going. Brink is what happens when the same team turns the same urgency toward a problem with a longer time horizon.

Jonathan Naccache
Jonathan Naccache, PhD
Co-founder & CEO

“Gene therapy has been ‘almost there’ for thirty years. The biology works. The delivery works. Precise insertion is the last unsolved step, and it’s exactly the kind of specific, hard problem we’re built to attack.”

Harry Kemble
Harry Kemble, PhD
Co-founder & CTO

“I’ve long been driven to find ways of generating data at a scale that makes previously intractable problems tractable. At Brink, we’re doing exactly that for one of the hardest protein design challenges in biotech.”

Andrew Griffiths
Prof. Andrew Griffiths
Co-founder

“Directed evolution is the most powerful design force we know of. We have spent decades building the tools to harness it at unprecedented scale. Brink is the cleanest application of that work I have ever seen.

Careers

Join us.

We're a team of 13 at the frontier of directed evolution, metagenomics, and protein AI, and we're growing. If you want to work on one of the hardest and most impactful problems in biotech, we'd like to hear from you.

See open positions →
Scientific Advisory Board

Expanding; further advisors to be announced in 2026.

Marshall Stark
Marshall Stark
Prof. Emeritus, Univ. Glasgow, Structure & Mechanism of Large Serine Recombinases
“Most groups pick one or two LSRs and push them as far as they’ll go. Brink is operating at a completely different order of magnitude.”
Justin Eyquem
Justin Eyquem
Associate Professor, UCSF, Pioneer in CAR-T cell engineering and site-specific genome integration
“Programmable recombinases represent a fundamental shift in genome engineering, moving from cutting DNA to precisely rewriting it with a simple, programmable system. The ability to install large, defined payloads at specific loci in both dividing and non-dividing cells opens new possibilities for immune cell engineering, enabling precise control over cell identity, function, and persistence. More broadly, this technology has the potential to become a foundational platform for in vivo gene and cell therapies.”