Vansh Dhiman

I’m an AI Engineer who ships production LLM agents end to end. I built and deployed a live autonomous lead-response agent that qualifies and scores inbound real-estate leads in under 60 seconds using Groq llama-3.3-70b, alongside the full-stack React platform it runs on.

As a Co-founder & AI Engineer, I own end-to-end AI-agent engineering—from LLM orchestration to reliable RAG pipelines. I designed a RAG-style LLM pipeline with structured prompt engineering, retrieval-augmented context injection, and output guardrails to support HOT/WARM/COLD lead scoring, and I integrated multi-step automation workflows through REST-based service architecture in Python and Node.js.

I also shipped the production web platform (auth flow, tiered pricing, demo-booking, and a live AI-agent dashboard) built with React + Vite and deployed on Vercel with a Git pull-request workflow and a CI/CD build pipeline. Before that, I did a Cybersecurity Industrial Trainee stint where I ran vulnerability assessments and penetration tests using Kali Linux, strengthening my focus on IAM, threat detection, and log/forensic thinking.