I'm a Pre final-year student who got deep into Generative AI — and never looked back.

My current obsession is RAG (Retrieval-Augmented Generation) — building systems that let LLMs reason over real documents instead of hallucinating. I've built a full RAG pipeline from scratch: multi-format document ingestion (PDF, Markdown, CSV), semantic chunking, sentence-transformer embeddings, ChromaDB + FAISS vector storage, and Groq LLM API for grounded generation.

What excites me most is the full loop: ingest → chunk → embed → retrieve → generate

Beyond RAG, I've built:

→ An autonomous AI agent using Google Gemini 2.0 Flash with custom prompt engineering, Selenium-based automation, and production-grade safety filters

→ An Explainable IDS combining Random Forest/XGBoost with a RAG pipeline for natural-language threat explanations

→ A full-stack AI chat app (React + Gemini API) shipped solo and deployed on Vercel

My background in cybersecurity gives me an edge — I think about LLM systems with a security and reliability mindset: prompt injection, data leakage, safe fallbacks, and auditability.

I'm actively building, learning, and looking for Generative AI / AI Engineering internship opportunities where I can contribute to real products.

📍 Bengaluru | 🔗 github.com/tej949