Service Commitments

This page describes how the service is actually operated today. We would rather publish a short list of true commitments than a long list of aspirational ones. Contractual availability targets and service credits are agreed per customer during onboarding, not published here.

1. How the service is run

• The platform runs on Google Cloud Run in London (europe-west2), in a single region. There is no multi-region failover today; a regional outage would make the platform unavailable until the region recovers or we redeploy elsewhere.
• Every deploy is gated: the candidate revision receives no traffic until an automated smoke suite of customer journeys passes against it.
• An automated probe exercises the production journeys every 30 minutes; failures page the team.
• Live component health is visible to anyone at /status.html. We do not yet publish historical uptime figures, and we will not quote numbers we do not measure.

2. Incidents

Service-affecting incidents are recorded at /incidents.html. Customers with an active engagement are notified by email. There is no public phone or 24/7 support line at this time.

3. Maintenance

Deploys use Cloud Run revision cutover and do not normally interrupt service. Any maintenance expected to interrupt service will be announced on the status page in advance.

4. Data processing

Data Processing Agreement: /dpa.html · Sub-processors: /sub-processors.html · Data residency: /data-residency.html (primary data is stored in Google Cloud, London, europe-west2).

5. Contact

Service enquiries and incident reports: team@neuravant.ai