Jeevansh Bhatia

I’m a B.Tech in Computer Science & Engineering student focused on building AI products with Retrieval-Augmented Generation and deploying end-to-end pipelines. I like combining strong fundamentals with hands-on experimentation—especially where I can iterate quickly and learn fast.

In my RAG-Based Network Anomaly Detection project, I built an embedding and semantic search pipeline for NSL-KDD, using sentence embeddings and similarity scoring to retrieve relevant historical records. I engineered the retrieval layer with ChromaDB, prompted a Qwen2.5-7B-Instruct model with retrieved context, and deployed the system as an interactive Gradio app, including a side-by-side evaluation to show improved reasoning quality.

I also build practical systems: an IoT-based food spoilage detection prototype using sensor data and ML logic, and an IoT smart monitoring pipeline with real-time transfer/processing and MongoDB storage. Alongside this, I develop Android apps in Kotlin with clean UI logic and backend integration, and I’ve won an IoT hackathon organized by my department and Cosmic Attire.