By product
Sovereign AI Workspace13 tools, one login · try free Sovereign AI ServersSealed on-prem appliance Sigmix Guard™PCIe scrubbing & audit firmware Sigmix ZeroSovereign Windows 11 endpoint StocksIQThe AI analyst for the Saudi market Sigmix GrowthPilotCross-platform CAC & budget AI Muraqaba AISeparate product · video AI Sigmix RedAI penetration testing · newBy industry
Banking & FinanceSAMA CSF · on-prem Public SectorNCA ECC-2 · sealed HealthcarePDPL · sealed desktops EducationLighter, adaptive tone EnterpriseCustomer-facing suite · CRMSigmix Guard™
Our PCIe card, running Sigmix-authored firmware — scrubbing & audit, in dedicated silicon.
Explore Sigmix Guard
Sovereignty & compliance
Security postureNCA ECC-2 · SAMA CSF · PDPL · ZATCA Zero-egress architectureDefault-deny · canary-monitored Sigmix Guard™Independent enforcement layerFor developers
API & DocsSovereign, compatible API System StatusAll systems operational Trust CenterAudit evidence & certsNothing leaves your site
Every request served on-box, sealed from the internet.
See your postureAll systems operational
In-Kingdom services · updated continuously
Ninety days, per service.
These are the services Sigmix operates on your behalf. Sealed on-prem and on-device deployments are not listed here — see below for why that is deliberate.
Where the hosted services run.
Every region on this page is inside Saudi Arabia. Nothing here fails over across a border — if a region is degraded, traffic moves to another in-Kingdom region or it waits.
Primary compute and storage for the KSA Cloud postures.
Secondary in-Kingdom capacity and regional failover.
Additional in-Kingdom capacity for the eastern province.
We can't see it. That's the product working.
A sealed appliance sends us no telemetry at all.
Most vendors monitor your deployment and show it on a status page. We cannot, because a zero-egress appliance has no outbound path — and adding one so we could watch it would quietly undo the thing you bought. The absence of your appliance from this page is the guarantee, not a gap in it.
Health, GPU utilisation, model readiness and the egress counter are all visible to you, in your own admin console, on the box. If you want that fed into your own monitoring, the appliance exposes /v1/health on your network and the audit anchors replicate to your SIEM.
What has gone wrong, and when.
Every incident affecting a hosted service is posted here with a timeline and, once resolved, a plain-language cause. We do not quietly delete them afterwards.
No incidents in the last 90 days
The most recent 90 days are clear across all hosted services. Full history is retained for 12 months.
// scheduled maintenance is announced at least 5 business days in advance, outside Kingdom business hours where possible
Get told before you notice.
Subscribe to status updates.
Incident notifications and scheduled-maintenance notices, in Arabic and English. Available by email or webhook, so you can pipe them straight into your own operations channel.
Request status alerts