I've been working on (and using) an open-source project called MooseStack that makes adding powerful real-time analytics to your apps feel like writing regular application code.

Most apps use OLTP databases (Postgres, MySQL, etc.) for transactions. But when you need dashboards, user behavior insights, aggregations over millions of rows, or real-time metrics, you hit a wall. Setting up a proper OLAP system (ClickHouse is excellent for this) traditionally involves complex infrastructure, Kafka/Redpanda setups, materialized views, schema migrations, and separate APIs. It often feels disconnected from normal dev workflows.

MooseStack is a developer framework (TypeScript + Python) that sits on top of:

ClickHouse (super-fast columnar OLAP)

Redpanda (streaming)

Temporal (reliable workflows)

The killer feature: moose dev spins up the entire local stack with hot reload. You define schemas, ingestion, materialized views, and APIs as code in your IDE. Change a file → everything updates instantly. It's also built as an agent harness, so it plays very nicely with AI coding tools (Cursor, Claude, etc.).

It lowers the barrier for full-stack or solo devs to add production-grade analytics without becoming data engineers.

Wanna check it out?

Would love to know your feedback, they do have Slack Channel, if you have similar requirements you can reachout to the team on thier slack.