Back to projects

Case study

Zapspace

Zapspace is a personal product I built end to end: one workspace for notes, audio, images, and video, then tasks and AI agents that work from what you've captured. Capture an idea the moment it hits, connect it to follow-ups, and act in the same space.

Visit live site

The Challenge (The “Why”)

Ideas arrive as text, voice, or media, but follow-through usually lives somewhere else. Notes apps, task lists, and AI tools stayed disconnected, so context got lost and agents had no coherent brain built from what users had already captured. Students, professionals, and executives needed one place to go from spark to done without juggling tabs.

The Technical Strategy (The “How”)

I designed and implemented the full architecture myself on TypeScript with TanStack Start and Nitro as the server runtime. The product model centers on capture (rich notes and Markdown, voice, images, and video), connect (linking notes to tasks and agents so follow-ups keep their context), and act (finishing work in the same workspace, including research that agents turn into executable tasks).

The Key Features & Contributions

  • Owned end-to-end product architecture and implementation for Zapspace
  • Notes workspace with rich text and Markdown so ideas are captured the moment they appear
  • Media and voice capture so audio, images, and video live beside notes in one space
  • Tasks linked to the notes they came from, keeping deadlines in context
  • AI agent connections that read and manage the workspace as a shared brain
  • Research-to-tasks flows so exploration becomes a plan users can execute

The Impact (The “Results”)

Zapspace ships as a single capture-to-action workspace: notes, media, tasks, and agents in one product, aimed at how students, professionals, and executives actually work, including offline-friendly moments when context would otherwise scatter.

Technical Stack

TypeScript TanStack Start Nitro

© 2026 - Jeffrey Ango, Senior Frontend Engineer. Built with Astro, Tailwind CSS and deployed with Vercel.