Insights

Notes from the build.

Original writing on multi-specific design, cryptographic provenance, the honest quantum question, and what we’re learning building NOVA-3 in the open. New posts weekly.

ScienceJun 5, 2026·7 min

A plain-English primer on multi-specific antibodies

What bispecifics, trispecifics, and VHH single-domain binders actually are, why pharma keeps reaching for them, and why they are hard to build.

ThesisMay 30, 2026·7 min

Designed, not screened: why we compose antibodies instead of hunting for them

Screening treats antibody discovery as a search problem; we treat it as a composition problem, and that change reorganizes everything downstream.

Coming up · weekly

Jun 12, 2026Predict vs. compose: where folding models stop and design begins

Jun 19, 2026What pLDDT actually tells you (and what it doesn't)

Jun 26, 2026Evidence Bundles: notary-grade provenance for pharma diligence

Jul 3, 2026Our honest take on quantum in drug discovery

Jul 10, 2026The DBTL flywheel: why wet-lab data compounds

Jul 17, 2026Inside the Module Library: composing from validated parts

Jul 24, 2026T-cell engagers 101: bispecifics, trispecifics, and the exhaustion problem

Aug 7, 2026Why carbohydrate targets are hard (and where quantum helps)

Posts publish Fridays. To get notified, email insights@qailabs.co