Tacit enables frontier labs to develop LLM capabilities for the life sciences by collecting high quality data on how science is done.
We imagine a future in which AI is a powerful tool for scientific discovery, capable of helping us reason through complex problems and driving optimal decision-making across discovery and development pipelines.
Tacit collaborates with biotechs and academic labs to evaluate models on high-value, real-world tasks and works with frontier AI labs to systematically improve model performance on those tasks. Data collection will focus on outcome-verifiable, long-horizon planning and reasoning tasks to power the most critical functions in the life sciences.
Tacit is supported by Alana Goyal at Basecase, Terrence Rohan at Otherwise, Soleio, Thomas Wolf (Hugging Face), Jonathan Frankle (MosaicML), Matthew Leavitt (Datology), Dan Shipper (Every), and others.