AI Engineer · LLM Systems · Agentic AI · RAG Pipelines
- CS undergrad turning research into production-ready AI*
I'm a B.Tech CS + Data Science student who loves building AI systems that actually work in production — not just demos.
My focus is on Agentic AI, where I design multi-agent workflows, semantic RAG pipelines, and LLM-powered backends. I care about making AI systems reliable, explainable, and secure — not just impressive.
"Good AI isn't just smart — it's trustworthy."
🤖 Multi-agent orchestration systems with real-time reasoning
🔍 Semantic RAG pipelines with reranking & grounding verification
⚡ Streaming AI backends (SSE) with security guardrails
🔐 Production-safe LLM apps with JWT, encryption & structured logging
LLM & Agentic AI
Vector Search & Retrieval
Backend & APIs
ML & Data
Automating Indian civil dispute resolution — end to end
A production-grade system built around a 10-agent, 7-stage AI pipeline that handles legal dispute analysis, evidence evaluation, and AI-generated verdicts. Features semantic RAG with BGE embeddings, CrossEncoder reranking, bias detection, and SSE real-time streaming — with enterprise-level security throughout.
Stack: Python · FastAPI · React · PostgreSQL · LangChain · pgvector
Chat with any YouTube video — context-aware, memory-enabled
A conversational AI agent that transcribes, summarizes, and answers questions about YouTube videos with persistent memory across turns — no more scrubbing through long content.
Stack: Python · LangChain · Groq · Streamlit
- Advanced context engineering and LLM reasoning patterns
- Graph-based multi-agent architectures with LangGraph
- Evaluation frameworks for production RAG systems
- Scalable async AI backends at production load
Open to collaborations on anything at the intersection of AI reliability, agentic systems, and real-world impact.
Let's build something meaningful → vidhigupta1500@gmail.com