AI engineer and software builder working across intelligent systems, applied research, and developer tools.
A technical case study on multi-agent dermoscopic classification, exploring what happens when model disagreement is preserved as an explicit signal rather than averaged away into an argmax.
I built a disagreement-aware vision system with two models, uncertainty signals, spatial explanations, and a learned arbitrator. An arithmetic mean beat it. That turned out to be the useful result.
No additional work is available yet.
Currently authoring applied AI & systems research papers. Preprints and publications will be published here soon.
