B.E. student at BMS College of Engineering, Bengaluru, working across real-time multi-signal fusion, LLM agents, time-series forecasting, and autonomous robotics.
Sharanya Sandur — B.E. in Artificial Intelligence and Data Science at BMS College of Engineering, Bengaluru (2023–2027).
Marketing Head of the VarAInce Club, BMSCE's AI & Data Science department club, directing outreach and cross-functional promotion for its events and initiatives.
Builds end-to-end — from Bayesian signal fusion and LLM-grounded agents to time-series forecasting pipelines and ROS2 robotics simulations — and competes actively on hackathon circuits.
Real-time candidate identification system that fuses multiple weak signals — participant metadata, audio, and meeting behavior — into one continuously-updated confidence score via multi-signal Bayesian fusion, with the Anthropic API integrated for explainable candidate selection.
Quantum-assisted energy optimization for an autonomous lunar rover — simulated on ROS2 + Gazebo, fusing LiDAR, IMU, and encoder data for real-time terrain perception, with a modular mapping and path-planning architecture exploring QAOA-based quantum optimization for energy-efficient routing.
Terminal-based support triage agent built over a 700+ document markdown corpus, indexed from scratch with BM25. Detects prompt injection, out-of-scope requests, and high-risk escalation triggers before generating a grounded response via the Claude API.
Forecasts website traffic from Google Trends search data, benchmarking Facebook Prophet against an LSTM model and surfacing predictions through an interactive Streamlit dashboard for resource and marketing decisions.
Full-stack DDoS attack simulator: generates synthetic network traffic, visualizes it as an interactive graph, and produces natural-language attack analysis via LangChain, on a Next.js frontend backed by FastAPI.
CineSense — a movie recommender that blends content-based filtering with personalized star ratings, serving similar-title and personalized suggestions through a Flask API.