MEGHA UKKALI
I’m an AI Engineer and software developer with 6+ years of experience building and deploying RAG, agentic AI, and multi-agent systems using Python and LangChain/LangGraph.
About
I’m an AI Engineer and software developer with 6+ years of experience building and deploying RAG, agentic AI, and multi-agent systems using Python and LangChain/LangGraph. I enjoy designing end-to-end solutions—chunking and hybrid retrieval (BM25 + vectors), agent workflows with routing and validation, and grounded response generation with LLM evaluation and reliability controls.
I’ve worked across production environments, from microservices and distributed systems to CI/CD, AWS deployments, and observability/testing. Recently, I built an agentic assistant for operational users at A.P. Moller Maersk that answers natural-language questions from both SOP knowledge and structured SQL/container data, and I’ve also delivered full-stack services in financial domain applications at Chatham Financial while continuing to explore multi-turn RAG and multi-agent research pipelines.
- 6 years' experience
- Joined October 2026
Skills· 15
Languages
- EnglishAdvanced
Industry experience
- Software & Internet
- Financial Services
- Computers & Electronics
- Professional Services
Work experience· 5 roles
Built an agentic AI assistant for terminal operations that answers questions about operational procedures and structured container/vessel data using natural language. Implemented an SOP-driven RAG pipeline with chunking, embeddings, and hybrid BM25 + vector retrieval in OpenSearch, providing source-backed responses. Developed LangGraph routing workflows to classify queries into SQL, RAG, or hybrid paths, enabling retrieval from PostgreSQL and the RAG knowledge base. Added reliability controls including structured routing outputs, SQL validation, retrieval checks, and grounded response generation. Containerized and deployed a FastAPI service on AWS ECS Fargate, integrating Amazon ECR, RDS PostgreSQL, OpenSearch, S3, CloudWatch, and GitHub Actions CI/CD.
Developed and maintained services for financial risk management systems that support pricing and valuation of derivatives for capital markets clients. Optimized a bulk update workflow used in schedule filling, reducing execution time from 6s to 3s (50% improvement) using temporary tables. Built RESTful APIs for CRUD operations and integrated services for data transfer between applications. Wrote unit, integration, and acceptance tests using xUnit and NSubstitute. Implemented full-stack features with ASP.NET Core/.NET microservices, React and TypeScript, REST APIs, MS SQL Server, Redis, and RabbitMQ. Continued working remotely from India via Velocity Global after U.S. visa expiration.
Built backend applications using Java 1.8, Spring MVC, Spring Boot, microservices, and Hibernate with SQL and AWS services (EC2, S3, RDS, IAM, SNS). Containerized services using Docker and added testing with JUnit. Designed and deployed Spring REST web services on an AWS EC2 instance and used SNS for notifications, supported with unit tests in JUnit.
Education
Hire MEGHA UKKALI today
Post a job and invite MEGHA, or contact them directly. No hidden commission.