I'm a 3rd-year B.Tech student in Artificial Intelligence & Data Science at Dr. Mahalingam College of Engineering and Technology, Coimbatore (CGPA: 8.14).

I specialize in building production-grade RAG pipelines, LLM-powered chatbots, and vector search systems. My focus is on making AI systems accurate, fast, and hallucination-resistant — not just demos.

• 🏆 Smart India Hackathon 2025 Winner — Smart Education category | ₹1.5 Lakh prize
• 🤖 Built a Legal AI Chatbot with hybrid retrieval (Qdrant + BM25 + Knowledge Graph + CrossEncoder reranking)
• ⚡ Experienced with local LLM inference using Ollama and LangGraph pipelines
• 📊 Built an AI-powered analytics platform with Gemini API, auto-ML, and interactive dashboards
• 📱 Also build cross-platform apps with Flutter + Firebase

LLM & RAG

Backend & APIs

Mobile & Frontend

Databases & Tools

RAG chatbot for legal document Q&A with hybrid retrieval and answer verification

• Hybrid search: Qdrant (dense) + BM25 (sparse) + Knowledge Graph for maximum retrieval accuracy
• Pipeline: LangGraph + Ollama (local LLM) with CrossEncoder reranking
• Answer verification layer to minimize hallucinations and ensure high-confidence outputs
• FastAPI Qdrant BM25 LangGraph Ollama Python Docker

Production-ready RAG REST API with semantic filtering and structured JSON output

• Vector search via Qdrant with local LLM inference using Ollama
• Docker-containerized vector DB — zero external API dependency
• Structured JSON output with schema validation and filtering logic
• FastAPI Qdrant Ollama Docker Python

Centralized mobile platform for student activity submission and dashboard tracking

• Secure REST API integration with JWT-based authentication over HTTPS
• Real-time activity tracking dashboard built with Flutter
• Flutter REST APIs JWT Dart

End-to-end mobile delivery app with real-time sync and cloud image storage

• Firebase Authentication + Cloud Firestore for real-time data sync
• Cloudinary for scalable cloud-based image storage
• Flutter Firebase Cloudinary Dart

Reinforcement learning environment for automated machine learning

• RL agent optimizes model selection and hyperparameter tuning
• Built with FastAPI backend
• Python FastAPI Reinforcement Learning

• 🥇 Smart India Hackathon 2025 Winner — Smart Education | Team HackHounds | ₹1.5 Lakh
• 🎙️ Office Bearer, Invictus — AI & DS Association (organized technical events & seminars)

LangGraph Agents      →  multi-step reasoning pipelines
Advanced RAG          →  self-RAG, corrective RAG, adaptive retrieval
LLM Evaluation        →  RAGAS, faithfulness scoring
System Design         →  scalable AI backend architecture

Open to AI/LLM internships, research collaborations, and hackathons.
Let's build something that actually works.