Chai scales like enterprise software across pharma

Chai is proving that an AI drug discovery company can scale more like enterprise software than like a biotech pipeline. Instead of betting all its capital on a few internal programs, it sells access to models and workflow software into many pharma teams at once, which lets it spread the same design engine across Novartis, Bristol Myers Squibb, argenx, Pfizer, and Lilly while avoiding the cost and timeline of running trials itself.

• The practical advantage is parallelism. A pharma partner can put Chai inside an antibody or protein discovery workflow, have internal scientists use the model on their own targets, and pay for access or licenses, while Chai can support several programs at the same time instead of advancing one asset at a time.
  1 sacra 4 chaidiscovery
• That creates a very different risk profile from Xaira. Xaira is built around an internal R&D loop for designing its own therapeutics, which means more upside per successful drug, but also much higher capital needs and slower scaling because every program pulls management attention, lab work, and eventually clinical dollars.
  5 sacra
• A close comparable is Isomorphic Labs, which has also signed multiple large pharma collaborations. The difference is that Chai is described and staffed more like a deployable product platform, with platform engineers and forward deployed scientists, which fits a many customer, many workflow model better than a single company pipeline model.
  2 chaidiscovery 4 chaidiscovery 6 isomorphiclabs

This model pushes AI drug discovery toward a split market. A few companies will act like platform vendors that sell picks and shovels to many biopharma customers, and a smaller set like Xaira will use AI mainly to build proprietary drugs. If Chai keeps adding partners, its main asset becomes the installed base inside pharma discovery teams, not any one molecule.