Some systems degraded

8 of 10 components fully operational · last updated Jul 30, 2026, 11:40 AM UTC

99.96%

90-day uptime

Open incidents

Active

minorINC-2026-0714Started Jul 30, 2026, 09:12 AM UTC

Elevated latency on AI inference in eu-west

MonitoringJul 30, 2026, 11:40 AM UTC
Capacity has been added and p95 latency is back under 900ms. We are holding the incident open for another two hours to confirm it holds through peak.
IdentifiedJul 30, 2026, 10:05 AM UTC
A model routing pool in eu-west is running below target capacity after a provider-side node recycle. Requests are succeeding but p95 latency is roughly 3× normal. Generators and the assistant are affected; playbook execution is not.
InvestigatingJul 30, 2026, 09:12 AM UTC
We are investigating reports of slow responses from AI generators in eu-west.

maintenanceMNT-2026-0729Started Jul 30, 2026, 02:00 AM UTC

Scheduled maintenance — backup verification workers

In progressJul 30, 2026, 02:00 AM UTC
Rolling upgrade of the verification worker fleet. Scheduled backups continue to run; restore verification reports may be delayed by up to four hours. No customer action required.

Components

Last 90 days · each bar is one day

Platform

Control plane

Orchestration API, scheduling and the approval fabric.

84ms p9599.99%Operational

90 days agous-east · eu-west · ap-southToday

Execution runners

Scoped runners that execute playbooks against customer estates.

141ms p9599.97%Operational

90 days agous-east · eu-west · ap-southToday

AI inference

Model routing for generators, analysers and the assistant.

612ms p9599.82%Degraded performance

90 days agous-east · eu-westToday

Secrets broker

Short-lived credential issuance. No standing secrets held.

39ms p95100.00%Operational

90 days agous-east · eu-west · ap-southToday

Customer surfaces

Customer portal

Tickets, invoices, assets, documents and reporting.

212ms p9599.98%Operational

90 days agoglobal-edgeToday

Website & docs

Marketing site, documentation, knowledge base and marketplace.

68ms p95100.00%Operational

90 days agoglobal-edgeToday

Helpdesk & ticketing

Intake across email, portal, phone and chat.

174ms p9599.96%Operational

90 days agous-east · eu-westToday

Operations

Monitoring ingest

Telemetry pipeline feeding alerting and anomaly detection.

96ms p9599.95%Operational

90 days agous-east · eu-west · ap-southToday

Security operations

24×7 SOC tooling, SIEM correlation and response automation.

99.99%Operational

90 days agous-east · eu-westToday

Backup & recovery

Managed backup scheduling, verification and restore.

99.91%Under maintenance

90 days agous-east · eu-west · ap-southToday

Incident history

Subscribe to updates

INC-2026-0621majorJune 21, 2026 · resolved in 84 minutes

Portal login failures for Microsoft SSO tenants

Root cause was a certificate rotation on our SAML signing key that was not propagated to the secondary IdP endpoint. The rotation job now validates both endpoints before completing, and we have added synthetic login checks per identity provider. Full post-incident review published to affected customers.

INC-2026-0512minorMay 12, 2026 · resolved in 153 minutes

Delayed monitoring ingest in ap-south

Ingest backlog cleared. Alert evaluation was delayed by up to 11 minutes for ap-south estates during the window; no alerts were lost. Partition count has been increased and backlog alerting now fires at a lower threshold.

Every incident rated major or above gets a written post-incident review sent to affected customers within five working days, whether or not anyone asks for one.

Something wrong that is not listed here?

If you are seeing a problem this page says does not exist, that is the most useful message you can send us. Open a P1 and it goes straight to an on-call engineer.

Raise a ticket Contact support

Loading_

Active

minorINC-2026-0714Started Jul 30, 2026, 09:12 AM UTC

Elevated latency on AI inference in eu-west

MonitoringJul 30, 2026, 11:40 AM UTC
Capacity has been added and p95 latency is back under 900ms. We are holding the incident open for another two hours to confirm it holds through peak.
IdentifiedJul 30, 2026, 10:05 AM UTC
A model routing pool in eu-west is running below target capacity after a provider-side node recycle. Requests are succeeding but p95 latency is roughly 3× normal. Generators and the assistant are affected; playbook execution is not.
InvestigatingJul 30, 2026, 09:12 AM UTC
We are investigating reports of slow responses from AI generators in eu-west.

maintenanceMNT-2026-0729Started Jul 30, 2026, 02:00 AM UTC

Scheduled maintenance — backup verification workers

In progressJul 30, 2026, 02:00 AM UTC
Rolling upgrade of the verification worker fleet. Scheduled backups continue to run; restore verification reports may be delayed by up to four hours. No customer action required.