Klippa had 7 outages in the last 2 years totaling 49h 35m of downtime — averaging 0.3 incidents per month.
There were 7 Klippa outages since January 21, 2026 totaling 49h 35m of downtime. Each is summarised below — incident details, duration, and resolution information.
The app is sometimes unresponsive when trying to save or do an action on a transaction.
The issue seems resolved, we are monitoring closely.
The financial model is currently experiencing availability issues causing requests to time out. We are investigating the issue.
This issue has now been resolved. It was caused by newly added server capacity not picking up incoming requests quickly enough. We apologise for any inconvenience.
We identified and resolved an issue where newly provisioned GPU nodes could not load the drivers required for model inference. As a mitigation, workloads have been moved to unaffected hardware, and all services have been running stable for the past 30 minutes. We will continue to monitor the situation closely.
We’ve identified an issue on the Doxis AI.dp platform where flows could fail with a “Memory limit exceeded” error despite their memory usage being below the limit. We’ve adjusted the service’s deployment configuration and are actively monitoring the cluster to determine whether this fully resolves the issue.
We have identified the underlying issue and implemented a workaround to resolve it. We’ll monitor the cluster for further abnormalities.
We have identified the underlying issue and implemented a workaround to resolve it. We’ll monitor the cluster for further abnormalities.
The DocHorizon Flowbuilder is currently unavailable due to infrastructure issues. We are actively looking into it. Flows and webhooks are currently not being processed.
The issues with the flow builder have been resolved, flows and webhooks are operating as normal again.
Between 01 April 2026, 00:00 CET and 02 April 2026, 10:00 CET, our Peppol Access Point experienced a disruption that temporarily prevented the exchange of E-Invoices with certain other Peppol Access Points. The issue has been fully resolved, and normal service has resumed. We apologize for any inconvenience caused and appreciate your patience and understanding.
Requests made to the DocHorizon Prompt Builder service may fail due to a model deployment issue at our hosting provider Microsoft Azure. We recommend retrying requests and expect that the issue is resolved soon.
We’re seeing greatly increased success rates on prompt builder calls and are monitoring the situation. We are awaiting confirmation from Microsoft Azure that issues have been resolved before closing this incident.
Azure has indicated that the issue has been resolved.
We’re currently experiencing availability issues on our Legacy OCR API product due to unexpected high load on the database. We have found the root cause and are currently implementing a workaround.
A workaround has been implemented and the API is available again.