Council-based crop disease vision
A mixture-of-experts framework routing 298K+ images across crop-specialist CNNs and a confidence-gated vision-language expert.
96.14% accuracy · 84.00 Macro F1
Open to roles & graduate study · 2026
Computer Scientist · AI / ML Engineer
Raza Hamid
I build applied AI systems that hold up in the real world — from air-gapped retrieval to award-grade computer vision.
0%
Accuracy on a 155-class crop-disease benchmark
0×
Faster internal retrieval for a public-sector client
0
IEEE / ICONIP-track research publications
0
Engineering internships across AI, data & security
Selected highlights
Research-grade ideas, shipped as working systems.
Air-gapped RAG for government
A fully offline retrieval system — on-prem indexing, zero external calls, hybrid search and reranking — for confidential documents.
~10× faster retrievalPublished & peer-reviewed
Lead author on a council-based fusion paper accepted to the Digi-Health workshop at IEEE COMPSAC 2026, Madrid.
IEEE COMPSAC 2026 · ICONIP 2026
Let's build something
Looking for an engineer who ships ideas, not just trains models.
I'm open to NLP / LLM engineering, MLOps, software engineering and data roles — and to graduate research programs.