I'm an Agentic AI Engineer with hands on experience designing and shipping production multi agent LLM systems and RAG pipelines, deployed live with CI/CD. Built three end to end AI platforms independently | spanning vector ingestion, LangGraph state machine orchestration, RAGAS evaluated retrieval, streaming inference APIs, and React frontends using LangChain, LangGraph, FastAPI, and Groq. GATE 2026 DA (Data Science & AI) AIR 2515, placing in the top percentile nationally. Seeking to apply agentic AI engineering depth to production grade LLM systems at a fast moving AI team.

• 🤖 Core Focus: Agentic AI systems · Multi agent LLM orchestration · Corrective RAG pipelines
• 🎯 GATE 2026 AIR 2515 in Data Science & AI · AIR 5711 in CS & IT
• 🎓 9.44/10 CGPA | B.Tech CSE, RGUKT RK Valley (2023 to 2027)
• 🔭 Seeking roles as Agentic AI Engineer / AI Engineer / ML Engineer | internship or full time
• 🛠️ Building with LangChain, LangGraph, Groq, FastAPI, FAISS, React
• 🌱 Currently exploring multi agent architectures, tool augmented LLMs, and RAG evaluation

LangGraph · Corrective RAG · FAISS · CrossEncoder · Groq Vision · FastAPI · React · Redis · Prometheus

🔴 Live Demo · 📖 API Docs · 💻 GitHub

Engineered a self correcting Corrective RAG (CRAG) pipeline as a LangGraph StateGraph with conditional routing across retrieve, evaluate, web search fallback, synthesize, and reflect nodes. Deployed five specialized agentic pipelines (Crop Advisor, Market Analyst, Schemes Expert, Weather Analyst, Leaf Scanner) over async SSE streaming on FastAPI.

• 🔬 Two stage retrieval: FAISS vector search (paraphrase-multilingual-MiniLM-L12-v2) + CrossEncoder reranking
• 📊 RAGAS evaluated: 0.80 Faithfulness · 0.90 Answer Relevance · 0.70 Context Recall on a 30 pair evaluation set
• 👁️ Groq Vision (Llama 4 Scout 17B) for multimodal leaf disease diagnosis under 3s
• 🌐 LLM translation pipeline across 11 Indian languages
• ☁️ Deployed on Vercel + Render with GitHub Actions CI, Redis caching, and Prometheus observability

Chrome Manifest V3 · LangChain · FastAPI · MongoDB · Redis · Tavily · Web Speech API

🟢 Live API · 💻 GitHub

Built a Chrome Manifest V3 extension injecting a Shadow DOM isolated AI chat sidebar into any webpage, grounding LLM responses in up to 3,000 characters of live page context via a LangChain powered FastAPI backend with async MongoDB (Motor) persistence and JWT authentication, deployed live on Render.

• 🔍 Integrated Tavily Search API for real time highlight to ask web search
• ⚡ Redis backed session cache and rate limited authentication endpoints
• 🌐 In place page translation into 9 Indian regional languages via Groq LLMs with one click DOM restoration
• 🎙️ Full voice to inference to audio pipeline using the Web Speech API

LangGraph · Multi Agent · FastAPI · React · TypeScript · Groq · APScheduler

💻 GitHub

Designed a LangGraph state graph agent system with four specialized agents (General, Technology, Finance, Sports) using LLM based query classification for intelligent routing across Google News RSS and DuckDuckGo as dual retrieval sources.

• 📧 Autonomous scheduled digest generation with APScheduler driven email delivery
• 🌐 Multilingual translation into 19+ languages via Groq LLMs
• ⚙️ Served through a FastAPI backend and React/Tailwind frontend