Autonomy Realms
Public Commons Claim Realm Docs

Onboarding

This guide walks you through the essential first steps after claiming your realm.

Connect Your AI Accounts

Autonomy Realms uses external AI services for transcription, analysis, and embedding generation. You need to provide your own API keys.

OpenAI (required). Go to Settings and enter your OpenAI API key. This powers transcription (Whisper), analysis (GPT), and embedding generation (text-embedding-3-small). Without this key, none of the AI features will function.

Anthropic Claude (optional). If you prefer Claude for analysis, you can add an Anthropic API key in Settings. When configured, you can choose which model runs your analyses. Claude is only used for analysis — transcription and embeddings still use OpenAI.

Create Your First Signal

Once your AI accounts are connected:

  1. Navigate to Signals and click Create Signal.
  2. Choose a signal type — Transmission (video/audio recording), Document (written text), Photo (image), or Hyperlink (external URL).
  3. Fill in the required fields and upload your file.
  4. Save the signal.

Understanding the Processing Pipeline

After you save a signal, the platform runs a series of background jobs depending on the signal type:

  1. Transcription — For transmissions, audio is extracted and sent to Whisper for transcription.
  2. Thumbnail Generation — Video signals get an auto-generated thumbnail.
  3. Analysis — The AI reads the signal content and generates structured metadata (title, summary, tags, themes, and more).
  4. Embedding Generation — The signal's content is converted into a vector embedding for semantic search.

You can monitor processing status from the notification panel. Each step updates independently — transcription might complete while analysis is still running.

Next Steps

  • Analyze your signal — select which fields you want the AI to fill in and run analysis.
  • Create a cluster — group related signals into a thematic arc.
  • Explore the Atlas — if your signal has location data, find it on the map.
  • Configure templates — customize the prompts used for analysis and embedding in Settings.
  • Set up reflections — generate AI reflections on your signals for deeper interpretive layers.