Intelligence

The database platform that explains itself.

Monitoring produces signals. Intelligence turns them into an argument you can check — what is happening, why it may be happening, what needs attention, and what to investigate first.

Why it works here

Intelligence is only as good as its context.

A general-purpose assistant knows a great deal about databases and nothing about yours. SchemaPulse already holds the operational context of every instance you run — metrics, queries, topology, elections, incidents, backups, recovery points and inventory — because it needs that context to run the platform.

That is the whole difference. The intelligence layer is not reading documentation; it is reading your estate.

It is also one layer, not a shelf of separate features. Correlation feeds investigation; investigation feeds the assistant; the assistant and the recommendation engine cite the same evidence and reach into the same subsystems — monitoring,incidents,query analytics andbackup and recovery. Ask a recommendation a follow-up and you are talking to the assistant.

AI Database Assistantorders-galera · last 24h

What changed before this incident?

Three things changed in the 40 minutes before INC-2291 opened.

At 02:14 UTC orders-galera-1 lost and regained cluster membership. Write throughput shifted to galera-0, whose innodb_buffer_pool hit rate then fell from 99.2% to 94.1%.

At 02:31 a deployment added an index to orders.line_items, and one query changed plan — mean latency 8ms to 340ms.

The latency alert at 02:47 is most consistent with the plan change, not the membership event: the membership change resolved 12 minutes before latency began climbing.

Evidence

Electiongalera-1 · term 5 · 02:14:08Z
Metricbuffer_pool_hit_rate · 99.2% → 94.1%
Querydigest a3f19c · 8ms → 340ms mean
Deployschema change · 02:31:44Z
Fig. 1An assistant answer separates correlation from cause, and states which of two candidate explanations the evidence supports.