At HiVanya, I architected a cloud-native multimodal AI platform that ingests data from more than 15 SaaS tools. I built a production-grade vector RAG system using Bedrock, OpenSearch, ECS, and Lambda for timeline-aware search and insights.
I also implemented infrastructure as code, CI/CD, monitoring, and event-driven microservices for secure, fault-tolerant ingestion. I worked on a Chrome extension that improves prompting quality and routes queries to suitable LLMs.
At Foundry, IDC Canada, I engineered ETL and real-time ingestion pipelines with AWS and Snowflake. I also developed microservices and backend services for a B2B intent analytics platform, including daily intent-lead pipelines and search across account and contact data.
Earlier, at Cians Analytics, I developed data pipelines, scraped and cleansed public data, and automated pipeline-volume monitoring. At India Private Limited, I worked on in-house ERP development for a used car startup using Core PHP and web technologies.
