Building production AI products end-to-end — from architecture design through go-to-market. Shipped HireOS (LangGraph + Gemini 2.0 Flash, live), KundaliAI (TanStack Start + Supabase), and LaptopPulse (Python daemon). Managing full product lifecycle including AI product management, UX/UI, payments integration, and deployment.

Evaluated frontier LLM outputs across code generation and reasoning tasks for RLHF training pipelines. Designed multi-agent evaluation frameworks, benchmarked 4 frontier models across 5 capability dimensions, and developed structured failure taxonomies that informed annotation guidelines for subsequent training rounds.

Intensive self-directed learning in AI/ML, completing B.Tech (CSE) at AKTU, and building foundational skills in Python, machine learning, and software development. Built the expertise that would later enable frontier model evaluation and AI product development.

Multifaceted role covering quality assurance, production management, and team leadership for pre-engineered metal buildings. Applied systematic quality methodologies and led cross-functional teams in manufacturing environments.