Transfyr, an Observability Layer for Science

Field notes

The homepage describes a sensor stack and multimodal models that observe laboratory protocols at atomic-action detail, with the goal of preserving contextual knowledge usually left out of written procedures. Its stated applications include identifying which variables matter, improving SOPs and coaching, translating manual work into automation, and supporting imitation learning for robots. The page also reports a $25 million seed round led by General Catalyst and links to a New York Times feature about researchers’ “magic hands.”