← Developer Experience · Case studies Operations · Case study

Faster detection & resolution

Observability and AIOps that drive mean-time-to-detect and mean-time-to-recover down — Datadog and resolve.ai for detection, ServiceNow Now Assist for resolution — backed by error budgets, runbooks, and blameless postmortems.

Overview

Incidents were noticed late and resolved slowly, with on-call engineers piecing together context by hand. The aim was to cut MTTD (mean time to detect) and MTTR (mean time to recover) with AI-assisted operations across the holistic product lifecycle (PLDC).

Datadog and resolve.ai shorten detection and triage; ServiceNow Now Assist accelerates resolution. Around them sit error budgets, runbooks, and blameless postmortems that make the improvement durable, release over release.

The challenge

Problems were detected late — often after users felt them
On-call engineers reconstructed context manually under pressure
Resolution was slow and inconsistent across teams
Lessons from incidents were not reliably fed back in

Our approach

Instrumented services with Datadog for end-to-end observability — metrics, traces, and logs in one place
Used resolve.ai to assist detection and root-cause analysis, cutting mean time to detect
Brought in ServiceNow Now Assist to accelerate triage and resolution, cutting mean time to recover
Set error budgets and on-call practices, with runbooks for common failure modes
Ran blameless postmortems and fed fixes back into the platform and the golden path

Results & business impact

Problems are detected sooner — closer to the cause, before users feel them (MTTD ↓)
Resolution is faster and more consistent with AI-assisted triage (MTTR ↓)
On-call carries less cognitive load thanks to unified observability
Incident learnings improve the platform release over release

Tools & technology

Datadog resolve.ai ServiceNow Now Assist MTTD MTTR Observability Error budgets Runbooks SRE