Scripted requests run from outside your infrastructure on a schedule, the way you measure 'what does a user actually experience'.
Synthetic monitoring (sometimes called active monitoring or proactive monitoring) is the practice of running scripted requests against your production endpoints from external locations, typically every 1-5 minutes, and recording latency, success, and content correctness. The script can be a simple HTTP GET, a multi-step user flow (login, search, checkout), or a full headless-browser session. Because the requests come from outside your infrastructure, they catch failures real-user monitoring (RUM) misses: a regional DNS issue, a CDN outage, a TLS-cert expiry, a recently-broken third-party dependency.
Real-user monitoring tells you what already-loaded users experienced. Synthetic monitoring tells you what new users will experience right now. The combination is essential because synthetic catches incidents during low-traffic windows (overnight, weekends) when RUM has no signal, and it constrains your SLO measurement to a controlled, comparable workload.
See the part of the platform that handles synthetic monitoring in production.