Tapas Das

AI/ML Software Engineer | Advanced Techniques Enthusiast | M.Tech in Computer Science | Contributing to tech growth at Intel Corp

• Developed and deployed a production-grade Retrieval-Augmented Generation (RAG) system using Azure OpenAI, LangChain, and PostgreSQL, enhancing support assistant response accuracy by 30%.
• Designed and implemented a supervised ML pipeline to predict optimal CPU configurations, reducing job cycle time by ~40%.
• Engineered custom document loaders and markdown-based chunking strategies, improving large language model (LLM) response accuracy by 25–30% across diverse document types.
• Developed specialized SQL agents and loaders for unstructured design documents, achieving a 40% increase in parsing accuracy compared to traditional methods.
• Integrated cutting-edge technologies including LangChain Indexing, LOTR (Lord of the Retrievers), RAGAS, and LangChain Agents, contributing to a notable boost in retrieval precision and overall system performance.
• Proficiently managed multiple project environments, collaborating with cross-functional teams and maintaining 99.9% uptime with strict SLA adherence.
• Demonstrated theoretical and hands-on understanding of ANN, CNN, RNN, LSTM, transformers, and attention mechanisms, used in experimental and prototyping efforts.

• M.Tech, Computer Science and Engineering National Institute of Technology, Mizoram