Oriel — AI & technology worth noticing, every day.

How Circle K accelerated DevOps incident triage to under 15 minutes with Amazon Bedrock

Web page · https://aws.amazon.com/blogs/machine-learning/how-circle-k-accelerated-devops-incident-triage-to-under-15-minutes-with-amazon-bedrock

Circle K's DevOps team cut incident triage from 45-60 minutes to under 15 by building an agentic triage system on Amazon Bedrock. This post walks through the architecture, the initial Amazon Bedrock Flows implementation, the migration to Amazon Bedrock AgentCore, and the lessons learned from scaling it.

First seen by Oriel on .

AI summary

Circle K built an Amazon Bedrock-based incident triage system that investigates AWS, logs, traces, and code, reducing mean triage time from 45–60 minutes to under 15 minutes.

  • Grafana alerts flow through Amazon SNS, Amazon SQS, Lambda, and specialized Amazon Bedrock agents querying AWS resources, OpenSearch, GitHub, and a knowledge base.
  • The system correlates findings into a structured Slack summary containing probable root cause, evidence, affected service and owner, and recommended remediation steps.
  • Circle K migrated from Amazon Bedrock Flows to AgentCore Runtime as investigations grew, gaining sessions up to 8 hours, persistence, OpenTelemetry observability, and framework flexibility.
  • The AgentCore version deployed in April 2026 reduced mean triage time by 67–75 percent; straightforward incidents take 5–8 minutes and complex incidents finish within 15 minutes.
  • Circle K found that pre-querying logs, tuning retrieval and generation, and narrowing code searches using earlier findings improved speed and result quality.

#amazon bedrock #incident response #devops #llm agents

Generated by gpt-5.6-sol. Check the original for details.

Aliases

SourceExternal IDURL
AWS Machine Learning Blog2ed0532609086bac2f9175750e90dc9bbbd0ab45https://aws.amazon.com/blogs/machine-learning/how-circle-k-accelerated-devops-incident-triage-to-under-15-minutes-with-amazon-bedrock/ Read discussion