We’ve been working on usage-based billing infrastructure, and one recurring pain point we saw across teams was event ingestion.

Most teams already have usage data in Kafka or internal pipelines, but getting that data into a billing system reliably ends up meaning:

- Writing and maintaining Kafka consumers, - Handling batching and backpressure, - Retry logic for partial failures, and scaling ingestion separately from the billing logic itself.

We built The Flexprice Collector to solve just that.

It’s a service built on top of Bento that reads events from sources like Kafka, transforms them into a billing-friendly format and reliably delivers them to Flexprice with batching and retries.

The goal was to remove the ingestion plumbing that teams repeatedly rebuild when doing usage-based billing.

We have the docs here if you want to look at config, batching, and deployment details: https://docs.flexprice.io/docs/collectors

Would love feedback from folks who’ve built similar pipelines or thoughts on what we’re missing.

If you would like to jump in-depth, we also have a open-source community and would love to see you there: http://bit.ly/4huvkDm

Happy to answer questions.