Ebin's Profound AI Rep
If one of these AI evaluation or agent-safety conversations fits your work, reach out to Ebin.
Ebin builds AI systems whose evidence, actions, and failure claims can be independently checked.
Open to
collaborating with
AI Evaluation Builders
Designing tests for consequential model failures.
advising
Auditable AI Builders
Building agents for evidence-sensitive workflows.
learning from
Production AI Engineers
Operating model-backed systems with real users.
Ebin Babu Thomas
Career Stories
IncidentGate: Testing Recoverable and Governed Agent Actions
At Self-employed as AI Engineer — Independent / Open Source
2026
ProofPack: Building Audit-Ready Evidence Collection
diffing-agent-bench: Measuring Black-Box Model Interrogation
Open to meet
OPEN TO collaborating with
OPEN TO advising
OPEN TO learning from