Oncology targets. A range of cancer types. One engine.
Oncology targets in active design across a range of tumor types, validation-ready as academic medical center collaborations are finalized. Oncology is the proving ground because the biology is genomically legible, the unmet need is severe, and the same design loop generalises to every domain that follows.
Designed, not discovered.
Precision oncology across 20 cancer types
Cancer types under active design
Each tumor type is addressed through a genomically defined driver rather than a histology label — which is why one engine covers them all.
Proprietary immune activation technology
Most oncology drugs try to stop a tumor. This programme does something different: it instructs the patient's own immune system to do the work.
Targeted Cell Delivery
Engineered cells reach the tumor site
Immune System Activation
The patient's own defences are switched on
Precision Tumor Elimination
Cancer cells are identified and destroyed
The patient is the therapy
We program the patient's own immune system to identify and destroy cancer cells. The therapeutic effect is generated by the patient's biology, directed by our design.
Designed, not discovered
Targeting, activation, and specificity are designed by the same engine that produces our small molecules — with immunogenicity and manufacturability scored from the first candidate.
Built for solid tumors
The programme is aimed squarely at solid tumors, where checkpoint therapy alone has not delivered and where precision of activation matters most.
Inhibition asks a tumor to stop. Activation asks the patient's immune system to finish.
Know the answer before the first dose
Before designing anything new, determine whether something that already exists will work. A patient's variant profile is read against the space of approved and investigational agents to predict response, non-response, and adverse-reaction risk ahead of the first dose.
The right therapy, the first time.
One genome, four answers — before treatment begins
Correction over lifetime management
Metabolic, cardiovascular, and autoimmune conditions are today managed rather than resolved — a patient takes a drug for decades. These are large, genomically stratifiable populations where a design engine can target the driver instead of the symptom.
Correction, not management.
Designed molecules for sustained therapeutic effect
When turnaround is the constraint
Resistant infection is the cleanest demonstration that the platform is not an oncology company in disguise. The constraint here is turnaround, not target novelty — which makes it the sharpest test of how fast the engine can go from biology to a manufacturable candidate.
Design speed is the therapy.
Biology to manufacturable candidate, on the clock
Correct the defect at its source
For hereditary disease the defect is known, singular, and inherited — so the therapeutic goal is correction rather than suppression. The problem is not a pathway to suppress for the rest of a patient's life; it is a sequence that is wrong, and can be made right once.
One treatment, lifetime cure.
Permanent gene correction — one treatment, one cure
The hardest domain, taken last
Degenerative and neurological disease is deliberately last. It is the domain that most depends on an engine that has already compounded outcome data from every patient treated in the domains before it — which is precisely the advantage that cannot be bought.
Instruction, not inhibition.
Signalling modelled end to end, receptor to response