At Cognizant, I developed Python workflow tools to streamline operational and developer workflows. I also built an MCP server that retrieves organization-specific technical documentation and provides context-aware information for AI-assisted development.
On my self-initiated Hybrid RAG System with Neural Reranker project, I engineered a retrieval pipeline combining dense and sparse search. I used FAISS and a Sentence-Transformers cross-encoder reranker, then integrated prompt evaluation and response-generation workflows.
My self-initiated MCP-Based Document Retrieval & Text Analytics project uses TF-IDF and KMeans for topic categorization and intent-based retrieval. I also integrated Tesseract OCR to extract text and structured information from documents for downstream AI processing.

