Built and published an open source CLI (TypeScript) that provides persistent project context across AI coding sessions, solving the session-reset problem for agent-driven development workflows.

Designed a YAML-driven architecture spec and drift detection system that identifies when AI agents deviate from declared stack decisions, with append-only audit logs.

Implemented a Model Context Protocol (MCP) server enabling direct integration with tools like Cursor, allowing agents to read project context and write back decisions through a structured protocol.

Designed and built a full-stack AI helpdesk product using Next.js, PostgreSQL, Cloudflare Workers, and Slack API, featuring semantic intent routing via Gemini/GPT.

Conducted competitive analysis across 15+ tools and 6 industry verticals, leading to a strategic pivot informed by user research and market gap analysis.

Modeled unit economics, pricing strategy, and infrastructure costs to evaluate business viability as a solo founder.