Instacart has introduced Blueberry, an AI-assisted incident response system designed to help on-call engineers investigate production problems from Slack.
InfoQ reports that Blueberry combines AI agents, operational data, and historical incident knowledge to generate grounded root-cause hypotheses. The system uses parallel subagents, MCP integrations, and incident history, while keeping engineers in control of the investigation.
The practical value is speed under pressure. During an incident, engineers often need to pull logs, dashboards, recent deployments, alerts, and past fixes into one picture. An assistant that can gather and summarize that context can reduce the time spent searching before a human makes a decision.
The human-in-the-loop design is important. Incident response is full of uncertainty, and a plausible but wrong explanation can waste precious time. Blueberry’s reported approach reflects a broader pattern in production AI: the agent proposes, organizes, and connects evidence, but engineers remain responsible for judgment and remediation.