Experiment Drill-Down

Audio-to-Action Experiments

I was curious what would happen if I talked to my computer. A few late-night iterations later, audio notes started turning into useful action items.

Workflow Visual

Record a short idea, send it into n8n, let transcription and parsing do their thing, then kick out an action item that I can actually use.

Screenshots and GIF Placeholders

I captured a few moments from the experiment flow and left space for future clips.

Mini Demo

Add a test action item and watch it move through the little workflow.

Current state: idle

Key Features

• One-click sample action items so you can poke the flow without setup.
• A tiny animation that makes the workflow feel alive.
• Friendly status updates instead of robot-sounding system logs.

Tech Stack in Plain Language

Next.js runs the interface, n8n handles the handoffs, AssemblyAI does the transcript heavy lifting, and Vercel keeps it simple to ship and tweak.

Lessons Learned

I tried recording a short audio clip and watching it become an action item. Here's what happened: short prompts gave cleaner tasks, and a tiny review step prevented weird edge-case outputs.

Next Ideas

I want to test richer categories, confidence scoring, and optional notifications that explain why an action item was created.