How Sherlocks AI uses Temporal to orchestrate AI agents for incident resolution

Akshat Sandhaliya, CTO and co-founder of Sherlocks AI, has developed an AI-powered Site Reliability Engineering (SRE) platform that autonomously handles production incidents to reduce Mean Time to Resolution (MTTR) from hours to minutes. The platform deploys AI agents to investigate incidents by querying observability tools and tracing issues back to their root causes, thus mimicking a seasoned engineer's approach. The challenge lay not in developing the AI but in ensuring the reliability of the agents, which often run multiple tasks in parallel across several stages, including identification, investigation, root cause analysis, and remediation. Sherlocks AI overcame these challenges by using Temporal, which unifies four distinct workflow processes—Knowledge Graph construction, infrastructure scanning, AI agent investigations, and event ingestion—into a single reliable system. Temporal provides durable execution, retry policies, scheduling, and model agnosticism, allowing for seamless integration of various models based on task requirements. This approach eliminates operational overhead, enhances system reliability, and supports continuous agent improvements. The platform also incorporates human-in-the-loop mechanisms for complex cases and uses Temporal’s Scheduler for proactive anomaly detection, ensuring that agents can react to potential issues before they escalate into incidents.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.