Supreet S

AI Engineer

Cognizant

Nov 2025 - Present • 10 mos

Developed Generative AI applications and agentic workflows using Python, FastAPI, LangChain, and OpenAI APIs,automating document-based query workflows and reducing manual information retrieval effort by 40%. Designed end-to-end RAG pipelines processing 500+ PDF documents through text extraction, cleaning, chunking,embedding generation, and FAISS-based semantic search, enabling context-aware question answering. Implemented the retrieval-to-generation workflow by matching user queries with relevant document chunks and passing retrieved context to LLMs, improving response relevance by 25% and reducing irrelevant responses by 20%. Integrated AWS S3, SQL, and REST APIs/Microservices to automate document ingestion, storage, retrieval, and AI-service communication, reducing manual processing steps by 35%. Optimized prompt engineering, document chunking, embedding generation, and retrieval configurations through iterative testing, improving LLM response quality and consistency by 20%. Debugged and tested AI workflows, resolved 20+ application issues, and used Git for version control, contributing to reliable deployment and delivery of production-oriented AI/ML solutions.