Ashish Kumar User

I’m a final-year Computer Science UG building AI/LLM-powered systems, with hands-on experience in agentic workflows using LangGraph and advanced RAG pipelines. I focus on end-to-end architectures that move from modeling to reliable deployment.

As a Machine Learning Intern (Team Lead), I trained and optimized real-time computer vision workflows using OpenCV and YOLOv8 for object detection and image classification on custom datasets. I engineered inference pipelines that connected frame preprocessing, model inference, and post-processing for production deployment.

In my Enterprise AI Knowledge Copilot project, I built an LLM-powered copilot with LangGraph integrating RAG and MCP-based tools for ticketing, employee lookup, policy retrieval, and internal knowledge search. I also implemented a Self-RAG/CRAG verification pipeline to reduce hallucinations by grounding responses in retrieved context and tool outputs.

I’ve also delivered multimodal and surveillance-focused assistants, including a multimodal RAG-based e-commerce product assistant and an AI-powered violence & weapon detection system for CCTV. I’m motivated by building scalable, asynchronous tool-driven backends (e.g., FastMCP + PostgreSQL) and shipping practical AI systems that perform accurately.