Built Around a Solvable Problem

the problem

We bring together immunology, protein engineering, and machine learning to understand how the immune system actually responds to cancer — and to build multispecific medicines designed to keep doing that as disease changes.

the platform

Coding Bio was founded in Oxford to make the immune system programmable — not as a metaphor, but as a method. FORGE is that system. The pipeline is the proof.

our beliefs

• Technology built the wrong way around. Drug discovery generates data as a byproduct of development. The companies that changed their industries built data infrastructure first — then built products on top of it. FORGE is that infrastructure layer.
• Scale changes what is learnable. Running 50 million functional datapoints across a screening cycle is not an incremental improvement on running 500. It changes what models can learn, what candidates can be designed, and what failures can be avoided before they reach the clinic.
• Capital efficiency follows from measuring the right signal early. Most attrition in drug development is predictable in retrospect. Selecting on functional immune-cell data — not binding — means fewer of the wrong candidates advance. Less time. Less capital. Better outcomes.