/ˈɒn.kɔːr eɪ aɪ/
An independent research laboratory advancing machine reasoning for medicine and law.
EncoreAI exists to make machine reasoning trustworthy where it matters most. We study how models read, weigh, and cite evidence in two domains where a wrong answer is not an inconvenience but a harm: clinical care and the law.
Our researchers work on retrieval, calibration, and provenance, the unglamorous foundations that decide whether a system can be audited, challenged, and ultimately relied upon. We publish our methods, our benchmarks, and our negative results.
We believe the path to dependable AI in high-stakes institutions runs through deeply capable, deeply inspectable systems, not through scale alone.