Questions, answered.
What is AIopsy?
AIopsy is AI-native uptime, workflow & AI monitoring. It checks that your site, APIs, user flows — and your AI — actually work, and tells you why the moment they don't. It works outside-in: you point it at a URL, describe what to watch in plain English, and it proposes the checks, verifies them, and starts monitoring.
Do I need to install an SDK or agents?
No. AIopsy is no-SDK and no-agent. It monitors from the outside, so there's nothing to add to your codebase — and it can monitor sites whose code you don't own.
How is it different from Sentry or other error trackers?
Error trackers are inside-out: you instrument your code with an SDK to capture exceptions, stack traces and traces from real traffic. AIopsy is outside-in plus AI: it checks whether the experience works (workflows, pages, APIs, AI answers) without any SDK, and diagnoses why it broke. They complement each other — AIopsy is not a replacement for in-code error tracking or APM.
Can it monitor my AI / LLM features?
Yes. AIopsy can call your LLM, RAG, or agent endpoints and judge the answer for quality — required facts, citations, on-topic correctness — not just whether the endpoint returns a 200. A 200 doesn't mean the answer is right.
What is “page health”?
Page health loads a page in a real headless browser and fails it when the page is genuinely broken — a non-2xx response, a blank or failed render, error text, or an uncaught JavaScript error — even when the server returns HTTP 200. A 200 can hide a page that's broken in the browser.
Can it monitor sub-pages, not just the homepage?
Yes. AIopsy can discover a site's first-level pages from the homepage and monitor each one, so a broken /pricing or /checkout doesn't slip by while the homepage looks fine.
How does the AI root-cause analysis work?
When checks fail, AIopsy groups the symptoms into a single incident, separates facts from hypotheses, and gives a likely cause grounded in your stack — correlated with what just deployed and any upstream-provider outages. It also remembers past incidents, so recurring issues are recognized on sight.
Does it replace Datadog / APM?
No. APM and distributed tracing give deep in-code performance visibility from real traffic. AIopsy is the outside-in + AI view — synthetic checks of whether the product works, plus AI incident triage. Run it alongside your APM.
Can agencies use it for client sites?
Yes. AIopsy has an agency / client mode: group apps by client, with branded public status pages and per-client reports. Because it's no-SDK, you can monitor client sites without touching their code.
How fast is setup?
About a minute. Describe what to watch in plain English; AIopsy probes your live system, proposes the right monitors, runs the first checks to verify, and you approve. No forms, no SDK.
Is there a free plan?
Yes — AIopsy is free during the preview. Paid plans add browser workflows, AI answer-quality checks, the AI page-health judge, status pages, and agency features.
Is my data and credentials secure?
Credentials you provide for checks are encrypted at rest, requests are SSRF-guarded, and access is scoped to your organization. See the Privacy Policy for details.
Still curious? See how AIopsy compares.