At Debound, I developed AI-powered backend systems for workflow automation, using multi-agent architectures for document processing and decision support.
I built production-grade services with RAG pipelines and agent workflows that reduced manual analysis effort by 50%. I also designed queue-based distributed processing with rate limiting, retries, and asynchronous execution, improving reliability and reducing failures by 35%.
Using LangGraph and Node.js, I developed multi-agent systems that translated natural language into database queries and analytics dashboards, reducing query execution time by 60%.
Earlier at Debound, I engineered an Infrastructure Management Suite and created a unified GraphQL API layer with a modular Node.js backend. I also resolved more than 30 backend issues and optimized CRM workflows to reduce manual effort by 60%. I later worked as a Full Stack AI Engineer at Crownstack.
