vCon Server is a pipeline for vCon conversation records. It takes them in, runs them through configurable chains of links (transcription, AI analysis, tagging, routing, webhooks) and writes the results to one or more storage backends.
Documentation: start at docs/index.md. The Docs workflow publishes the same pages to https://vcon-dev.github.io/vcon-server/.
You need Docker with the Compose plugin. The quick start takes you from a fresh clone to a processed vCon. It sets three values in .env, writes a small config.yml, and starts the stack.
| Audience | Start here |
|---|---|
| New users | Getting Started |
| Operators | Installation, Configuration, Operations |
| Developers | Contributing, Extending, Reference |
- Chain-based processing: compose reusable links into pipelines driven by Redis queues
- Processing links: transcription, AI analysis, tagging, routing, webhooks, SCITT and DataTrails records
- Storage backends: PostgreSQL, MongoDB, S3, Elasticsearch, Milvus, Redis, SFTP and more
- Multi-worker scaling: parallel workers, per-worker concurrency and parallel storage writes
- External ingress: scoped API keys let third-party systems submit vCons to specific queues
- OpenTelemetry: built-in tracing and metrics export
The server carries lawful_basis attachments through the pipeline. It does not create, validate or enforce consent. See lawful basis.
With the stack running:
docker compose run --rm conserver pytestContributing covers local setup with uv.
MIT. See LICENSE.