Megha Ukkali

I've built enterprise backend systems at Maersk and Chatham Financial, and I now apply that foundation to reliable LLM-powered applications. At Maersk, I implemented features for the T-One Smart Flow Optimizer, coordinating terminal operations through event-driven microservices.

I built Kafka and RabbitMQ workflows that consume real-time IoT events and trigger container-to-vessel routing, while adding Elasticsearch, Kibana, and Grafana observability for scheduling failures. I also deployed microservices to Kubernetes with GitHub Actions CI/CD.

At Chatham Financial, I created and maintained services for valuing and pricing financial derivatives, optimized a schedule-filling bulk update from six seconds to three, and delivered full-stack features with ASP.NET Core, React, TypeScript, SQL Server, Redis, and RabbitMQ.

I've also built agentic RAG, multi-agent research, and multi-turn document assistant systems using Python, LangChain, LangGraph, FastAPI, OpenSearch, FAISS, RAGAS, Airflow, AWS, and Docker. I'm looking to build dependable AI products grounded in strong retrieval, orchestration, evaluation, and deployment practices.