System status

Automated checks run every 15 minutes: each vendor's public status feed is read, authenticated connectivity probes verify Conduit can actually reach the services it depends on, and the platform's own public health endpoint is probed. Component status is the worst of its inputs. There is no hand-editable status anywhere in the pipeline — nobody can paint this page green.

It means our own measurements found impairment: a direct check answered slower than 2.5 seconds, or a vendor declared a major platform outage while our checks still pass. “Outage” means our checks are failing outright. “Unverified” means we currently have no trustworthy data — shown honestly instead of assuming everything is fine. In the daily bars, red is reserved for sustained downtime (more than an hour's worth of failed checks in a day); a brief blip outlines the day in red, and every cell's tooltip carries the exact measured percentage either way.

Vendors publish global status pages covering their whole platform, and a “minor issue” there often affects a region or sub-service we don't touch. We measure our own slice directly with authenticated checks, so when a vendor reports a minor issue but our measurements pass, the component stays operational and the advisory is disclosed underneath — we report what we measure, in both directions.

The bars show only days we actually measured — public monitoring began on August 26, 2026. Days before monitoring existed are shown as “no data,” never backfilled or estimated. The rightmost cell is today: it fills in live as each 15-minute check lands, then becomes a permanent history row when the day ends (UTC). A 90-day uptime number appears only once 90 days have genuinely been recorded.

Two consecutive failed checks open an incident automatically: it appears on this page, the team is alerted immediately, and the recovery is recorded by the same automated checks. The incident history below is machine-written — an outage cannot be quietly left out.

Yes. An independent external monitor (Better Stack) probes Conduit from outside our infrastructure and its signal feeds the “Conduit platform” status on this page — an outside witness our own systems can't influence.

Use the contact page or email support@conduitapp.ai. This page tracks platform-wide health; account-specific issues (a forwarding change at your carrier, a paused assistant) won't appear here but the support team will see them with you.