Hey HN! I built Primatomic because setting up event sourcing infrastructure is a pain. You either stuff events into Postgres and hope for the best, or you're wiring together Kafka, projection services, and snapshot logic before you've written any business code.

Primatomic lets you append events to a log and deploy a pure fold function (compiled to WASM) that derives your read state. You query it over HTTP. The platform handles snapshotting, consistency (read-your-write via position tokens), and S3 event archival.

The WASM/WIT contract means your view logic is portable and testable outside the platform - write it in Rust, Go, Python, JS, anything that compiles to WASM.

I'd love feedback on the developer experience, docs, or anything that feels unclear. There's a free tier and I'm happy to hand out free credits if you want to test larger workloads.