from
I build AI systems that have to show their <work/>. Agents that test the tools they write, evals that don't grade themselves, and logs that say why each call was made.
ai engineer
top 6
of 1,700+ teams at ABB
’27
0101234567890+
012+
0123
Snap-shots
I’ve been
Building At
Shipped Coodra to npm as @coodra/cli. Project context loads when a coding session opens, decisions are recorded while the agent works, and a policy hook checks every file write and shell command.
Tested an LLM as a prior-authorization reviewer on 166 cases: 63% agreement with policy, and its stated confidence said nothing about accuracy. Also ranked which hospital workflow to build AI for first.
Built the evaluation pipeline that graded 1,500+ HackHustle 2.0 submissions, grew registrations from 1,600 to 4,500+, and ran five student tech communities from 400 to 600+ members.
Built Vigil, a Kubernetes observability engine that keeps facts, relationships and forecasts apart, and presented it to ABB leadership.
Solved a vision-language model problem statement in 24 hours at Next-Gen. Also 1st of 50+ at Gidy Techpreneur and 2nd of 60+ at Fusion for Future.
Presented research on context management for small on-prem language models: summarize and re-plan mid-document so a short context window doesn't wreck long inputs.
B.Tech in Artificial Intelligence and Machine Learning, graduating 2027. Ran a hands-on small language model fine-tuning workshop for 70+ students.