Launching today
BROKA
One platform for your messaging infrastructure.
5 followers
One platform for your messaging infrastructure.
5 followers
One self-hosted console for Kafka, Redis, RabbitMQ and Apache Artemis: one permission model for every team, production guardrails and a tamper-evident audit trail. Kafka is free in Community.
Free Options
Launch Team
Framer AI AgentsDesign and publish professional sites with AI
Promoted
Hi everyone, I'm Yusuf, one of the people building BROKA.
Inside one company, different teams often run different brokers with different tools: one team on Kafka with its own UI, another on the RabbitMQ management UI, a third on the Redis CLI. Each has its own logins and permissions, and nobody has one picture of which service uses which topic or queue, who owns it, or who changed what. The changes that hurt are usually the routine ones, like an offset reset on the wrong consumer group or a purge on the wrong queue.
BROKA puts those brokers in one self-hosted console:
One permission model: users, teams and scoped roles across every broker you connect
Guardrails on production: read-only or guarded environments, a reason for destructive operations, offset resets previewed before they run
A tamper-evident audit trail you can export and verify
It runs on your own infrastructure with Docker Compose; there is no hosted version
The Kafka edition (Community) is free and needs no account. Redis, RabbitMQ, Apache Artemis and Memcached, single sign-on, Message Search and data masking are in the Commercial edition, which is now on sale.
For teams, the part we would point to first is Applications (Commercial). An application is one of your services, with its owner and its team. You bind it to the topics, queues and keys it uses, each with a role and a reason, and the dependencies between services fall out of those bindings. Alert rules can be scoped to an application and notify that team's Slack or Microsoft Teams channel, and its owner and team can edit it without being administrators.
Also in Commercial: an MCP server, so an AI assistant can work with your brokers under the same rules as a person. It acts as the owner of an API token, with that person's permissions and every guardrail. Changes go through preview, plan and apply, a destructive tool asks for a reason, and everything it does lands on the same audit trail. It stays off until an administrator turns it on.
We put particular weight on the audit trail. Every change and every refusal, on every broker you connect, lands on one tamper-evident trail that you can export and verify with a tool that isn't BROKA. In the Commercial edition, reads can be recorded too, and the trail can be forwarded to Kafka or RabbitMQ.
What comes next depends on you. We are looking at Apache Ignite, Apache Cassandra and possibly Apache Pulsar, and the feedback here will decide which comes first. We are also building on Debezium: change data capture into Redis and RabbitMQ is on our public roadmap.
What would help us most: if you run more than one broker, tell us which part of operating them is the most painful today, and what you would expect a console like this to do that it doesn't yet.