About
Bishal Upadhyaya.
I started in neuroscience, recording electrical signals in living neural circuits. Now I build AI systems that ship. Research on one end, products people use on the other.
What I do
Data pipelines, models, infrastructure, and the product on top. The bar is that it feels effortless to the person using it, not just impressive in a benchmark table.
- Agents and LLM systems that survive production. Tool-calling with structured outputs, multi-model routing, evals in the loop, honest failure reporting. The demos on my home page run on these.
- Retrieval and model efficiency. Agentic RAG, hybrid and semantic search over vector stores, and models distilled without giving up recall. The fingerprinting model is 27.7M parameters doing work most teams throw 300M at.
- Infrastructure a CTO can sign off on. Terraform-managed AWS and GCP, least-privilege credentials, SOC2-minded logging, cost governance with receipts.
Current focus
I cofounded Vibeset Technologies. Three products: Curation for setlists, Cue for music synced to picture, Choon for fingerprinting and provenance.
Music is just where these systems ship today. The same spine went into early Alzheimer's detection at UCSF, and model-efficiency research at Thermo Fisher with NVIDIA and HPE. The domain changes. The rigour doesn't.