Company
Different systems.
The same problem.
The insight
Incomplete systems
Decades of infrastructure prove that execution failure rarely comes from bad logic. It comes from stale or incomplete records.
The shift
From answer to action
When AI agents act on their own, hallucination isn't an inconvenience. It's operational liability.
Where we start
High-consequence first
Legal workflows demand strict context sufficiency, privilege protection, and auditable proof chains.
Yen Kha
Founder & CEO
2026 100 Changemakers at MITYen has spent nearly two decades building backend, data, cloud, and distributed systems across startups and large enterprises.
While co-founding Access to Justice for All at Stanford Law School's CodeX, she worked on AI for legal claim viability, where a decision depended on whether the system had the right evidence and context behind it. Her work has been presented at ICAIL, the International Conference on Artificial Intelligence and Law, and she has been an invited speaker for Bloomberg Law.
That problem became the foundation for GammaLex: infrastructure that gives AI the trusted context it needs to make better decisions.
Advisors
Bruce CahanJDStanford Lecturer · Founder, Access to Justice for AllAdvises on legal AI & regulatory contextThirty-eight years in data science and financial technology. He keeps our legal work grounded in how claims and regulation actually operate.
Martin BronkMDStanford MedicineAdvises on clinical AI & safetyA practising surgeon and Stanford clinical faculty member. He tests our thinking against what safe, high-stakes decisions require.