https://github.com/dbos-inc/dbos-transact-golang
DBOS workflows make your programs durable by automatically checkpointing their state to Postgres. If your program crashes or fails, all workflows seamlessly resume from their last completed step when your program restarts. This durability makes workflows useful for solving many different problems, including:
- Operating an AI agent, or anything that connects to an unreliable or non-deterministic API.
- Building observable and fault-tolerant data pipelines.
- Orchestrating business processes so they seamlessly recover from any failure.
- Running reliable background jobs with no timeouts.
- Processing incoming events (e.g. from Kafka) exactly once
What’s unique about DBOS is that it's just a Go package--all you need to add durable workflows to your app is included in this library (plus Postgres). There's no separate service or orchestrator to run, so you can incrementally add it to an existing Go application without rearchitecting anything.
Also, because it’s all Postgres-backed, you get all the tooling you’re familiar with: backups, GUIs, CLI tools–it all just works. You can use Supabase, RDS, Neon, or any other Postgres provider.
If you want to try it out, check out the quickstart:
https://docs.dbos.dev/quickstart
We'd love to hear what you think! We’ll be in the comments for the rest of the day to answer any questions.