At Edulyt India, I contributed to “AI-Driven Insights for Predicting Customer Purchase Behavior,” a project focused on using predictive models to analyze customer data and forecast buying patterns. I worked on data preprocessing and model training to generate actionable insights.
I designed and developed a modular, multi-agent framework to audit large language models across safety dimensions including prompt injection, toxicity, bias, and output safety. My Python audit orchestrator schedules specialized agents and aggregates their results into composite risk scores.
For a full-stack Cricket Information System developed during my Persistent Systems mentorship, I used Flask and AWS services including Cognito and DocumentDB. I integrated real-time sports data and news APIs and implemented authentication and subscriptions.
My other projects include an LSTM rainfall prediction model and web applications built with Flask and MongoDB. I also contributed to Linux monitoring automation as a tech intern at MobileComm Technologies (India) Pvt. Ltd.
