At TCS, I architected and developed a modular, multi-MCP AI platform for manufacturing intelligence. I built Meta MCP as a shared knowledge and data-access layer and developed domain-specific MCP servers for Advisory Analytics and Asset Twin workflows.
I implemented RAG pipelines with ChromaDB, vector embeddings, and semantic search, improving data-retrieval performance from 14 seconds to 1 second. I also designed AI services with caching, parallel execution, and asynchronous processing, reducing response time by 50%.

