Saroj Patil is an AI Researcher at SISA Information Security in Bengaluru, working on the Prism AI Security Suite — systems that discover, red-team, monitor, and govern AI across enterprise codebases. The work sits at the messy intersection of applied AI and security: shadow-AI and MCP discovery, AI supply-chain risk, agent analysis, and AIBOM generation.
Before SISA, Saroj interned at ISRO's Liquid Propulsion Systems Centre on spacecraft anomaly detection using autoencoders and diffusion models, and holds a B.E. in AI & Data Science (CGPA 9.26, 6th rank) from Global Academy of Technology. Two IEEE-published papers span the range of that curiosity — from dementia prognostication with computational models to asteroid-hazard assessment using explainable ML (ExtraTrees + SHAP).
Outside the day job, Saroj builds ML pipelines for astronomy — currently a gravitational-lens detector trained on simulated survey data — and creates AI-education content for India's developer community. An active Toastmasters member, Saroj enjoys turning dense technical ideas into talks people actually want to sit through.