AI Engineer with 4+ years of end-to-end experience building and deploying production-grade LLM, RAG, and deep learning systems. Adept at translating complex AI research into business impact.
Bridging the gap between AI research and production reality.
From fine-tuning LLaMA for domain-specific healthcare tasks to architecting multi-agent Graph RAG systems, I focus on building AI that is observable, scalable, and impactful.
Designing resilient multi-agent systems using LangGraph, Neo4j, and Vector Stores.
Multi-agent stock research system. LangGraph orchestrates agents across live market data, SEC 10-K filings, and news feeds. The quant layer handles all numeric computation in Python; agents interpret results and never do the math themselves, which keeps financial reasoning grounded and auditable.
Literature discovery on arXiv breaks when queries ignore citation structure and shared tasks. Research Pod pairs LangGraph agents with Neo4j relationship retrieval and Qdrant semantic chunks so answers trace both graph hops and dense similarity. Right now the focus is hardening evals and observability for production-style traceability.