Gaurav GULATI

Engineered robust Retrieval-Augmented Generation (RAG) pipelines for enterprise search, document processing, and LLM-powered automation.

Integrated fast vector databases with scalable Redis to enable hybrid search and low-latency generative responses.

Deployed modular APIs using FastAPI and constructed dynamic LLM chains with LangChain for tool calling and prompt management.

Developed components for processing unstructured data, utilizing embedding models, metadata filters, and custom retrieval mechanisms.

Engineered a full-stack Video Transcript Summariser and Audio Summariser pipeline leveraging the Gemini API, encompassing data extraction, chunking, summarization, and UI integration.

Built modular experimentation pipelines with OpenAI, Anthropic Claude, Gemini, and Hugging Face Transformers to evaluate LLM behavior and output quality.

Collaborated on developing LLM-driven automations for real-world applications including content generation, document understanding, and prompt-based task chaining.

Successfully completed a 2-month Generative AI internship at IGDTUW in partnership with Sansoftech Services Pvt. Ltd.