A fresh installation has an empty database. Configuring an AI provider or a live property provider does not populate the market datasets used by research tools.
Enable an ingestion scope
Automatic ingestion is opt-in. Set INGESTION_AUTOMATION_SCOPE_PROFILE_KEY on
the worker, then configure the sources and coverage required by that scope.
Without an active scope, the worker can still process onboarding jobs.
The application documents local-state-48 for development and production-us
for production ingestion tooling. A scope does not activate every source:
source credentials, readiness checks, and publication controls still apply.
Preserve source artifacts
The production Compose stack mounts the same persistent artifact volume in the API and worker. Preserve it along with PostgreSQL and Redis when updating.
Local artifact storage is the self-hosted default. S3-compatible shared storage is also supported; some national production ingestion operations require it. Configure the store for the scope and operation you intend to run.
Inspect available commands
The application image includes ingestion operator commands. List them without opening database or provider connections:
docker compose --env-file .env.production -f compose.production.yaml run --rm --no-deps worker ingest --help
Individual commands retain their existing plan, approval, budget, and scope checks. Read the command's help before running it against a deployment. A command's plan mode can still connect to the database.
Verify outcomes
Worker readiness is separate from successful data publication. Check ingestion evidence, source coverage, publication status, and dataset freshness for the sources you enable.
The application ingestion architecture guide describes source validation, publication, monitoring, and recovery controls.