- Services
- Kafka and Confluent
- Governance and center of excellence
Governance and center of excellence
Every topic needs an owner. Every schema needs a rule.
We set up schema governance, topic ownership, access policy and the standards your teams follow, then build the center of excellence that keeps them in force after we leave.
Updated October 19, 2026
- ChangeNew field type on the payment event
- Schema checkNot compatible with the registered version. Stopped.
- Topic ownerPayments platform lead notified
- Consumers affectedNone. The change never reached them.
- Decision loggedRecorded with time and reason
Where you are
Kafka spread through your company the way good tools do: one team, then several. Each cluster has its own naming, its own retention and its own idea of who may read what. Schemas live in application code, and the topic everyone depends on has no owner on paper.
What is in the way
Standards written in a wiki do not stop a bad deploy. Governance that holds is enforced by the platform: a registry that rejects an incompatible change, access that follows ownership, and a small group with the authority to say no. Most estates have the tools and not the authority, so we set up both together.
The same field change, twice.
No schema rule
Three consumers down, found downstream.
Registry with a rule
Stopped at the producer. Every consumer unaffected.
How we clear it
We make the platform enforce the rules, then hand the authority to a group your teams already trust.
- 01
Schema governance.
Every event schema in the registry with a compatibility rule, so a breaking change is stopped at the producer.
You hold: a registry with versioned schemas and the rule each one follows.
- 02
Topic ownership and access policy.
Naming, retention and access standards, a named owner for every topic, and access granted by policy instead of by request in a chat thread.
You hold: a topic inventory with an owner and an access rule for every entry.
- 03
A center of excellence your teams run.
Standards, a review process for new topics and schema changes, templates and training, run with your engineers until they own it.
You hold: a working center of excellence with its charter, its named reviewers and its review process already in use.
Questions we are asked about governance.
How do you govern Kafka topics and schemas?
With rules the platform enforces, not a document people are asked to follow. Schemas carry compatibility rules in the registry, every topic has an owner, and access follows that ownership.
What does an event streaming center of excellence do?
It sets the standards, reviews new topics and schema changes, and owns the platform roadmap. Your engineers staff it; we work alongside them at the start and step back once the reviews run without us.
Can AI agents read our events safely?
Yes. An agent is treated as a consumer like any other: it gets its own identity, reads only the topics its owner approved, and every request it makes is logged.
Also in Kafka and Confluent
Tell us which change broke production last time.
Talk to an architectTell us what you are working on. Our Customer Partnership Team will reply to arrange a conversation.
