Trackt — AI-Powered Follow-Through System

Trackt is a founder-designed and founder-built AI-assisted operational system for professionals who manage relationships through communication. It connects to email, scans active threads, and surfaces broken promises, missed follow-ups, unanswered questions, and approaching deadlines before they become relationship damage. It is framed as a risk dashboard, not a to-do list.

Most professionals spend the bulk of their day on communication tasks that should never have required their attention in the first place. Leads slip through cracks, commitments made in email get forgotten, and unanswered questions sit in threads for days. Follow-through lapses quietly erode client trust and revenue.

I conceived the product strategy, information architecture, risk-dashboard model, AI-detection workflow, and user experience, then built a working frontend MVP end to end. The dashboard is organized into three urgency tiers (At Risk / Needs Attention / Waiting on Others), each item carries a High/Medium/Low confidence score, and every flagged item includes a contextual suggested reply so detection closes to action in one screen.

Founder, product strategist, UI/UX designer, and developer. Owned the concept through architecture and a production-quality working MVP.

React, TypeScript, Hono, Tailwind CSS, Vite, Cloudflare (D1 + R2) architecture, and Claude for the AI analysis pipeline.

Working MVP is live as a functional frontend prototype with a full mock-data layer. The AI analysis engine, Gmail OAuth, and real data layer are the next build phases. Product waitlist and positioning live at usetrackt.app.

Product strategy, AI-assisted workflow design, full-stack web development, risk-UI design, and founder-level execution from concept to working build.