- reflexd replaces the Datadog agent: CPU, memory, disk, network, PHP-FPM, MySQL, Redis, and queue depth — all collected by default.
- Alert thresholds, rate alerts, and correlation groups replace Datadog monitors — configure per-server or via server groups.
- Deploy markers correlate releases with metric shifts, giving you Datadog-style APM context without a separate vendor.
- Brain playbooks go beyond alerting: when metrics breach thresholds, Reflex can repair autonomously with a full audit trail.
See Reflex fix production — not just page you
PagerDuty would call you right now. Reflex completes detect → diagnose → fix → verify — often in under 30 seconds.
Interactive walkthrough: trigger a php-fpm failure and watch the playbook repair it.
Open the sandbox demoThe 3am PHP-FPM crash
502s appear → Brain diagnoses worker exhaustion → restart and verify in under 30 seconds.
Disk filling up
Trajectory alert at 92% → proactive log rotation before customers hit errors.
OOM cascade
Kernel kill → dependent services restarted in safe order with incident summary.
After you sign in, open the importer at (Integrations → DeployHQ import).