I work on making AI agents verifiable, easier to monitor, and safer to trust. I run this research under Cygnux Labs, my independent lab for post-AGI infrastructure.
Previously, I was a Research Assistant at the Cambridge Judge Business School and worked at Microsoft on Azure core infrastructure. I studied Computer Science at ITER alongside Data Science at IIT Madras.
Current work:
- Proof-Gated-Signing: solver-checked transaction guards for onchain AI agents that still hold when the chain changes between check and execution.
- Tracekit: tamper-evident tracing, policy gates and cross-checks for AI coding agents. See what the agent did, and why.
- Frontier-Autolab: a 16-agent startup reinvents itself across 1990–2040, scored against real history, then used for forecasting.
Earlier work: multi-agent crypto trading (Onchain-AI-Hedge-Fund), AVX2-accelerated NTT for homomorphic encryption (HE-poly-core), WGAN-A2C portfolio optimization (GenAI-Portfolio-Management), and quantum secure direct communication (QSDC).
I'm always happy to talk about agent oversight, AI evals, verifiable infrastructure, and ambitious engineering problems. The best way to reach me is by email (bravish.ghosh_at_outlook.com).