Rishabh Didwania

I'm building production AI applications at MAQ Software, including a full-stack analytics product that uses NL2SQL query generation to reduce response times from 10 seconds to under 3 seconds.

I've modernized an enterprise conversational AI platform serving 2K+ monthly users by refactoring 10K+ lines of JavaScript and implementing caching, state management, and multi-turn workflows. I also maintain Node.js and Express services for crawler automation, QnA training, validation pipelines, and content ingestion.

My projects focus on grounded, reliable GenAI systems: RAG pipelines for semantic search across 2K+ policy documents, enterprise knowledge assistants using Azure AI Search, and multi-source research agents with tool calling and task orchestration.

Previously, I built a scalable lead management platform at Sense Original Technologies, improving lead-tracking efficiency by 30%. I enjoy improving retrieval quality, latency, and answer reliability across Python, FastAPI, React, PostgreSQL, GraphQL, and vector search.