At Agivant Technologies, I designed a knowledge graph schema for turning enterprise knowledge-base documents into an atomic, consistent, and maintainable representation. I also architected a pipeline to extract and structure information from PDF sources stored in Azure Blob Storage.
I evaluated Graph RAG against Hybrid RAG for enterprise retrieval and help-desk use cases, and defined an approach for incremental updates and consistency validation.
At Valutics Inc., I developed and maintained production AI systems spanning LLM-powered workflows, knowledge retrieval, and backend services. I improved knowledge graph and retrieval pipelines using Python, Neo4j, Redis, and vector search technologies.
I also contributed to agentic AI orchestration and resolved more than 25 production issues across AI pipelines, ETL workflows, and backend integrations. In my projects, I built a multi-agent customer support system with LangGraph and Azure AI Search, plus a natural-language-to-SQL system for querying SQL Server tables.
