Farsight

Farsight was born out of a deeply personal realization: oncologists are drowning in data. Every day, they navigate an overwhelming flood of clinical trials, genomic reports, imaging results, and treatment protocols — all while making life-or-death decisions for their patients. The idea was simple but ambitious: what if AI could serve as a trusted co-pilot, helping clinicians cut through the noise and focus on what matters most?

The project began at Jori Health, where the team was already building AI-native healthcare infrastructure. Farsight started as an internal tool — a way to surface relevant clinical trial matches and treatment insights for oncology cases. But as the prototype took shape, it became clear this was something much bigger. Oncologists who saw early demos described it as the tool they had been waiting for their entire careers.

Building Farsight meant tackling some of the hardest problems in health-tech. The system needed to ingest and reason over unstructured clinical notes, structured lab data, genomic sequencing results, and the latest published research — all in real time. The AI pipeline was designed to be explainable at every step, because in oncology, a black-box recommendation is not just unhelpful, it is dangerous.

The technical architecture combines large language models with retrieval-augmented generation over a continuously updated medical knowledge base. A custom embedding pipeline indexes clinical trials, FDA guidelines, and peer-reviewed oncology literature, while a separate reasoning layer maps patient-specific data against this knowledge graph.

Farsight represents a belief that AI in healthcare should augment human expertise, never replace it. The co-pilot metaphor is intentional — the oncologist is always in command, but now they have an intelligent partner scanning the horizon, spotting opportunities, and flagging risks that might otherwise be missed.