How Property Finder automated incident management with AWS DevOps Agent
Imagine receiving an alert about a production issue, having it automatically analyzed, a ticket created, the on-call engineer paged, and a fix deployed—all before most teams would finish their morning coffee. That’s the reality at Property Finder, a major property marketplace, which built an end-to-end autonomous incident management system using AWS DevOps Agent. What once took 2-3 days from initial detection to resolution now completes in 14 minutes. This isn’t science fiction; it’s what’s possible when you combine observability, automation, and AI-powered agents working together.
At its core, Property Finder’s system operates as an autonomous workflow: when CloudWatch detects an anomaly, AWS DevOps Agent springs into action as your tireless on-call engineer. The agent connects to your existing tools—CloudWatch Logs, metrics, and diagnostic APIs—to perform root cause analysis automatically. It doesn’t just flag the problem; it investigates it. Is the database slow? The agent checks query performance and connection pools. Is a service throwing errors? It examines logs, traces, and error rates to pinpoint the exact issue. Once it identifies the likely cause, the system creates a Jira ticket with detailed findings, pages the on-call engineer through PagerDuty, and opens a pull request with a remediation script—all without human intervention. This is intelligent automation: the agent handles the repetitive diagnostic work humans shouldn’t be doing at 2 AM.
What makes this approach powerful is that it dramatically reduces mean time to detection (MTTD) and mean time to recovery (MTTR). For teams managing microservices, this matters enormously. In a typical incident, engineers spend hours gathering logs, cross-referencing metrics, and narrowing down which component failed. Property Finder’s system eliminates that triage phase. By the time an engineer wakes up or reaches their desk, they’re not starting from scratch—they’re inheriting a pre-diagnosed problem with a proposed fix ready for review. This is especially valuable for incident patterns the system has learned from before; over time, the agent improves at recognizing and responding to similar failures. It’s also safer than you might think: the system pages humans and waits for approval before executing critical changes, maintaining human oversight while removing the manual grunt work.
For any organization running cloud infrastructure—particularly those with complex distributed systems—this pattern offers a tangible path forward. You don’t need to rebuild everything; you can start by automating the diagnostic phase for your highest-impact alerts, or focus on a single critical service. Connect your monitoring tools to an agent, define what “normal” looks like in your system, and let it learn your incident patterns. The value isn’t just in speed; it’s in consistency, reduced cognitive load during crises, and the ability to scale your on-call operations without hiring more engineers. As AI-powered agents mature, this kind of autonomous incident management is becoming table stakes for teams serious about reliability.