Category
External Dependency Intelligence
Independently observed, timestamped knowledge about the third-party services your infrastructure depends on - the record that answers “was it us or them?”
Definition
External Dependency Intelligence is the continuous observation of third-party APIs and services - from outside both your stack and the vendor’s - turned into attributable, timestamped records of behavior. Status pages describe intent; intelligence describes behavior.
Why it matters
Modern products are compositions of vendors: payments, auth, messaging, AI inference, cloud primitives. A monitoring system that watches only your own services has a blind spot shaped exactly like your dependency graph. When checkout fails, the alert says checkout is failing - not whether the cause is your deploy or the provider’s outage.
How it differs from uptime monitoring
- Subject: uptime monitoring watches your services; dependency intelligence watches your vendors.
- Origin: probes run from independent regions on infrastructure the vendor does not control.
- Verdict: quorum across regions, not a single failed request.
- Artifact: checksummed fault reports, not screenshots.
How RELIASTRA produces it
Fixed-interval checks per dependency per region, every result stored with region and timestamp, quorum-confirmed incidents, deterministic correlation against your own incident history, and verifiable SLA evidence. The full method is published in how RELIASTRA measures vendor reliability - including the cases we deliberately refuse to call an outage.
Practical example
At 14:02 your error rate spikes. Regional probes show the payment API timing out from two regions over the same window while your database latency is flat. The investigation starts at the vendor, the rollback is cancelled, and the credit conversation opens with a timestamped record instead of a theory.
What to do next
See it live on Track, understand dependency monitoring, or how the platform fits together.
Know when your dependencies fail. Prove what happened.
Monitor the external APIs your product relies on, correlate their failures with your incidents, and generate verifiable evidence when a vendor causes downtime.