Skip to main content

The Big Picture

How all the pieces of AutoPod fit together — from your live stream to a listener's headphones.

The five stages

Everything AutoPod does follows a simple pipeline:

1. Record

AutoPod continuously records your station's audio stream. It doesn't just record during scheduled shows — it records everything, 24/7. Audio is captured in small segments (approximately 40 seconds each) and stored securely.

This continuous recording is what makes the 60-day archive possible — even if a show wasn't in the schedule, the audio is there.

2. Match

When a scheduled show ends, AutoPod checks the broadcast schedule to determine:

  • Which show just aired
  • What time it started and ended
  • Which audio segments correspond to that time window

3. Stitch

The relevant audio segments are downloaded, concatenated in order, and encoded into a single MP3 file. Metadata is embedded — title, show name, artwork, and timestamps.

If you've configured a pre-roll or sponsor tag, it's prepended to the audio at this stage.

4. Check

The completed episode goes through a quality check. AutoPod validates that the episode duration is within expected bounds (matching the scheduled duration). If something went wrong — a missing segment, an unexpected gap — the system can retry automatically.

5. Distribute

The verified episode is published to all your distribution channels simultaneously:

RSS feeds are regenerated every 2 minutes, so podcast app subscribers receive new episodes shortly after they're published.

Key design decisions

Why record everything, not just scheduled shows? Because schedules change, shows overrun, and sometimes you want to publish something you didn't plan. The continuous recording means you always have the audio — the schedule just tells AutoPod how to slice it up.

Why separate recording from episode generation? Because they're different concerns. The recorder just captures audio reliably. The generator uses the schedule to make editorial decisions — what to publish, with what title, under which show. This separation means you can change your schedule, fix mistakes, and create episodes retroactively without losing audio.

Why store audio as segments? Small segments are resilient. If your stream drops for 40 seconds, you lose one segment instead of an entire recording. And segments can be shared across episodes — if two shows air back-to-back with no gap, the boundary segment is used by both.

Want to go deeper?