Prashant Jha

I’m an LLM Engineer and AI Engineer with 5 years of experience, transitioning from data science roots into full-stack LLM engineering. I focus on production-ready LLM pipelines—especially Retrieval Augmented Generation (RAG) and generative AI systems that behave consistently in the real world.

In my current work, I built an end-to-end AI job aggregation pipeline collecting 150+ daily remote AI/ML job postings with multi-region and multi-query coverage. I implemented a two-stage semantic filtering approach using embeddings and structured evaluation (OpenAI GPT models), reducing raw listings to ~30 highly relevant results.

I also designed evaluation-focused systems—like golden dataset generation for Copilot features and a document-grounding pipeline that creates reduced summaries and query-specific evidence views from enterprise DOCX/PPTX/PDF content. I’ve delivered multimodal extraction for visually rich documents, improved throughput with multithreaded batch execution and caching, and earned recognition such as the Winner of Smart India Hackathon 2020, a Stroke of Excellence Award 2023, and a published research paper on automated traffic management.