Lakshmi Meyyappan

I’m an AI Engineer with hands-on experience building, deploying, and debugging production-grade Generative AI and Agentic AI systems. I built InvoiceIQ AI, a live SaaS deployed on AWS that’s used by real users daily.

My work focuses on agent reliability and evaluation-driven iteration. I designed DocuMind AI, a 5-agent LangGraph orchestration system with planning loops, conditional state transitions, and human-in-the-loop checkpoints, and I fixed coordination failures by redesigning state schema to prevent looping.

I also engineer robust RAG pipelines. I created EduRAG to address silent retrieval failures by moving from pure dense retrieval to a hybrid FAISS + BM25 approach with reranking and confidence-based filtering.

I move confidently across the full stack—Python, FastAPI, React.js, and production cloud deployments on AWS—while choosing classical ML when it beats LLM reasoning. I’ve previously done 2 years of software engineering foundation work at Photon Infotech, and I bring that same end-to-end discipline to AI product design.