Increased latency in PROD US instance API

Incident Report for Sardine AI

Postmortem

Summary

Due to a database performance issue, API calls to /v1/customer , /v1/issuing/risks and /v1/feedbacks in June 25th faced increased latency.

What happened

One of the database we use to load certain customer and transaction data experienced slow latency for queries used in our customer and feedback APIs, causing latency issues on those APIs from 15:20 UTC to 16:23 UTC at June 25th. We've identified the root cause and are working on fixes to mitigate the issue.

Why it happened

One of our clients has very high volume and a particular usage pattern that didn't match well with our current database design for this data, causing the entire cluster to degrade in performance.

What We’re Doing About It

We are actively working on a solution to this issue, which includes moving the affected customer/transaction data to more performant databases, eliminating the risk altogether. In parallel we are also looking into actions to gracefully handle the usage patterns that cause degradation in the entire cluster.

Posted Jun 30, 2026 - 18:51 UTC

Resolved

The incident regarding the latency increased has been resolved.

It started at 15:20 UTC and lasting until 16:23 UTC.

We'll include a postmortem into this status page in the future.
Posted Jun 25, 2026 - 16:23 UTC

Investigating

We are currently investigating this issue.
Posted Jun 25, 2026 - 15:20 UTC
This incident affected: Customer APIs.