ProjectDrop

The problem

Project ideas, decisions, fixes, and next steps often get scattered across chats, documents, voice notes, and random notes. ProjectDrop gives that context somewhere structured to go.

The idea

Create a mobile-first project memory app where I can create a project folder, record a voice memo or add text, save it to the project, and generate useful AI outputs from the accumulated context.

What it does

• Creates project folders
• Captures text notes
• Records voice memos
• Uses API-based transcription
• Saves notes to the relevant project
• Generates summaries, action steps, reports, and coding-agent handoffs
• Connects into my wider workflow by letting me copy outputs on my phone and pull them into FlowDock on my Mac

Tech stack

• React Native
• TypeScript
• Mistral transcription
• DeepSeek AI outputs
• Backend details to be finalized before public release

What I built

• A mobile-first capture flow for project folders, voice memos, and text notes.
• Structured AI output flows for summaries, action steps, reports, and coding-agent handoffs.
• A workflow connection where generated handoffs can move from phone to Mac through FlowDock.

What I learned

• Useful AI features need structure, constraints, and clear output formats.
• The coding-agent handoff feature changed the app from simple note capture into a project memory system.

Challenges

• Refining API limits and AI output formats before wider testing.
• Keeping the app focused on useful project context instead of becoming another loose notes tool.

What I would improve next

• Finalize backend details before public release.
• Keep refining tester-ready output formats and limits.

Links