Harsh Vardhan

I’m an AI-focused intern who blends product support engineering with applied ML work. I diagnose user issues, troubleshoot platform workflows, and collaborate to reproduce, document, and resolve customer-reported bugs across the product lifecycle.

On the AI side, I designed and built a core scoring engine for an AI-powered career-matching platform. I integrated multi-dimensional quantitative signals to rank and surface best-fit candidates at scale, while engineering Python automation pipelines that enforce data integrity and remove manual bottlenecks.

I also build agentic systems and production-ready models. I created a production-ready multi-agent deep research assistant using Python, LangGraph, and the OpenAI API, with retrieval powered by real-time web search via Serper.dev for citation-backed reporting.

Across projects, I prioritize reliability, measurable performance, and deployability—optimizing models for production (including TensorRT), delivering inference through Flask REST APIs, and iterating toward accuracy and consistency. I’m motivated by teams that value careful engineering, fast learning, and user-impactful outcomes.