Gene Editing Hasn't Scaled.
Until Now.

The promise of gene editing has been trapped by slow, costly, and fragmented R&D.
We're building AI-native research infrastructure that makes scalable breakthroughs possible.

๐Ÿ† Recent Win · April 2026

Mandrake Bio won the GEM ร— Adaptyv RBX1 Binder Design Competition

1 of 322
Only "Strong"-classified binder in the entire competition
26 nM
Binding affinity (KD) to RBX1 โ€” strongest in the field
9 / 321
Designs were confirmed binders. We had the strongest.

Mandrake submitted an 11.8 kDa miniprotein binder generated by our ORBIT framework (Oracle-Reseeded Binder design with Interface Targeting). It achieved the competition's only "Strong" classification โ€” beating 321 other designs from labs around the world.

The Assay

Deep dives from the lab notebook โ€” protein engineering, enzyme design, and scaling gene editing.

Subscribe to The Assay

The Future

We're not building one gene editor. We're building the infrastructure to create thousands, making gene editing as scalable as software.

I would like to:

Join us to make life programmable. See open roles.

View Careers โ†’