Sanjana Padukone

I developed backend services for NavMind AI at Inventeron Technologies LLP, supporting intelligent traffic prediction and smart route recommendations.

I built FastAPI REST APIs integrating routing and weather data for real-time traffic analysis and ETA prediction, reaching 90–95ms average response latency. I also trained and evaluated supervised machine-learning models for congestion prediction, achieving up to 95% accuracy.

My projects include AutoCTI, a real-time AutoML cyber-threat intelligence system that detects phishing, malware, and brute-force attacks through automated classification, monitoring, and incident-response workflows. I also created a React.js dashboard with WebSockets for live threat visualization and alerting.

I built an AI Smart Router Agent using LangGraph, LangChain, Google Gemini, ChromaDB, RAG, and MCP to route queries between semantic document retrieval and tool-execution workflows. My AutoCTI research received a KSCST Research Grant and was published at IEEE AIDE 2026.