Skip to main content

​
What is mAItion?

mAItion is an AI knowledge platform that ingests data from connectors and lets you chat with that data.

​
Do I need OpenAI or OpenRouter to start?

Not for basic ingestion and retrieval. For rephrase/chat model calls, configure provider keys in env files.

​
Why do I not see my documents in answers yet?

Most common causes:
  • connector is not configured
  • ingestion schedule did not run yet
  • source credentials are wrong
Check connector env vars, config.yaml, and service health with docker compose ps.

​
How often does data refresh?

Per connector schedule. Configure schedule env vars (in seconds), for example:
  • S3_ACCOUNT1_SCHEDULES
  • MEDIAWIKI1_SCHEDULES
  • SERPAPI1_SCHEDULES

​
Can I use multiple connectors at once?

Yes. You can enable multiple source types and multiple instances of the same type in config.yaml.

​
Can I use multiple S3 accounts or buckets?

Yes. Add additional source entries and matching env vars (S3_ACCOUNT2_*, S3_ACCOUNT3_*, etc). You can also configure individual connectors to use different buckets to keep provenance separate.

​
What is the difference between frontend and backend env files?

  • .env.openwebui.example: Frontend runtime and admin bootstrap settings
  • .env.rag.example: RAG backend API/workers, connectors, rate limits, provider config

​
Is ENABLE_SIGNUP required?

Keep ENABLE_SIGNUP=True on first start for admin auto-provisioning, then disable later for single-user/private mode.

​
How can I improve answer quality?

  • Keep sources fresh (shorter schedules)
  • Use reliable embedding and inference models