Shubham Thakur

I’m a results-driven AI engineer who builds and deploys intelligent systems with Large Language Models (LLMs), AI agents, and agentic workflows. I translate complex AI ideas into scalable solutions that improve accuracy, consistency, and real user outcomes.

In my current role as an AI Prompt Engineer, I built a RAG-based AI assistant for an LMS using LangChain, OpenAI embeddings, and ChromaDB—enabling context-aware, multi-turn querying of course content. I’ve also engineered structured prompt frameworks (chain-of-thought, few-shot, system-role) that improved LLM output accuracy and consistency by 40%+, and delivered end-to-end LLM pipelines for automated content generation, quiz creation, and learner personalization for 10,000+ learners.

I’m equally focused on production readiness: I’ve deployed FastAPI microservices for AI REST endpoints, embedded streaming AI experiences into React/Next.js interfaces, and automated supporting workflows with n8n, Zapier, and Make.com. Previously, I optimized LLM prompts to reduce hallucinations by 45%, and I still enjoy building clean, user-friendly web interfaces—starting from an internship foundation in HTML5, CSS3, JavaScript, and React.js.