We've built robust data tracking systems before, for example at companies that hadn’t implemented an event-sourced architecture from the get-go and needed to scale with traditional transactional databases. We’ve since learned that the tech is useful to lots of others. We’ve gone full-time on Bemi now and have been following that demand to build a general-purpose tracking solution that can be used for a wide range of use cases such as reliable audit trails, surfacing data to customer support, reverting changes made in an API request, or testing different application states.

We’ve built Bemi to be lightweight and secure. It takes a practical approach to achieve the benefits of event sourcing without requiring rearchitecting existing code, switching to highly specialized databases, or using unnecessary git-like data versioning abstractions. We want your system to work the way it already does with PostgreSQL to allow keeping things as simple as possible.

We plug in at both the database and application levels, to get reliability, performance, and also a comprehensive understanding of every change.

On the database level, we ingest changes using change data capture (CDC). On the application level, our library allows passing application context and metadata to the write-ahead logs (WAL) automatically. We then stitch the change data together and store it in a structured format in a destination cloud PostgreSQL.

We plan to charge for compute if you’re a company storing >1M changes a month. If this is something that seems interesting to you feel free to play around with it or message us! We’re still early and it’d be amazing to get HN’s feedback.

What do you think HN?