Taktile had 14 outages in the last 2 years totaling 32h 9m of downtime — averaging 0.6 incidents per month.
There were 14 Taktile outages since September 4, 2024 totaling 32h 9m of downtime. Each is summarised below — incident details, duration, and resolution information.
Customers are experiencing intermittent errors when executing decisions. We've are investigating the issue
All systems operating as normal
We've noticed an issue with accessing Taktile's Frontend on app.taktile.com. We have found the issue and working on a fix. This does not impact production usage of Taktile.
Fix validated
Some customers may be unable to load the Decision History UI. The team has identified the issue and is rolling out the fix The API is unaffected and continues to work as expected.
We are observing intermittent failures for requests to Gemini models. This may impact automated decision flows and integrations.
The service is fully restored
[Investigating] Since ~ 18:00 UTC we are tracking intermittent API 5xx responses in us-east-1
We have not detected any new incidents in the last 30 minutes. It appears the underlying issue with AWS has been resolved; however, we are continuing to monitor the situation closely to ensure stability across all affected regions.
Seems to be an AWS incident, on-call team is working on it.
[Recovered] between 6:52am 9:22am UTC Oct 20th, all US Taktile customers were impacted by a major AWS outage in the US East (N. Virginia) Region. Taktile APIs returned 5xx error codes. As per our current understanding no other regions were affected. We will conduct Post Mortem analysis to understand how we could mitigate this better.
Some customers are experiencing login issues. Our team is investigating the cause and working on restoring access as soon as possible
The login issue has been resolved
Our team is currently investigating a failure for jobs to start.
Fix was deployed.
Engineering team is aware of this issue and we're currently working on mitigation.
All workspaces are now healthy.
We are observing elevated error levels on the Connections API. We've identified the root cause and will release a fix shortly
The incident is fully resolved; the fix has been rolled out to all customer workspaces.
We've observed that customers that use Azure AD cannot authenticate into the application right now. We're investigating the issue right now.
The issue has been resolved.
AWS S3 has reported an outage in us-east-2
AWS solved the issue on their end, all service should be operating normally from now on.
Test runs in the app are not working at the moment. We will release a fix shortly.
test-runs are operating normally
On Sept 4th, between 16:45 and 17:30 UTC, a small set of customers experienced sporadic failures calling our APIs. The root cause was an issue in the underlying AWS infrastructure.
Recovery was confirmed by AWS and we have notified affected customers directly.