Problem

Relvy AI addresses the operational burden of production incidents: when an on-call alert fires, engineers must sift through logs, metrics, traces, infrastructure, code changes, and prior incident context under time pressure. That creates costly interruptions, context switching, and dependence on the most experienced engineers; Relvy’s stated benefit is to reduce that on-call burden and give teams more time to build.

The killer use case is autonomous investigation of alerts from an on-call workflow, especially alerts already arriving in Slack from tools such as PagerDuty or Datadog. Instead of merely notifying an engineer, Relvy investigates the incident and produces a root-cause analysis, with the company claiming that 70% of alerts are resolved in under five minutes.

Product / Service

Relvy is an AI on-call engineer delivered as a software platform that connects to a company’s observability stack, code, infrastructure, and custom APIs. It uses continuously updated context from natural-language runbooks, prior investigations, and live telemetry to investigate alerts with purpose-built tools for logs, metrics, APM traces, and code. The resulting investigation notebook is auditable, visual, and reviewable by the engineering team rather than being an opaque chatbot response.

The workflow can be initiated from Slack, automatically for configured alerting bots, or through a mention or prompt. Teams can import or write plain-text runbooks to constrain and guide investigations, while Relvy’s forward-deployed engineers tune the system to the customer’s stack. The product supports both a hosted workflow and self-hosting in the customer’s own infrastructure, with integrations spanning observability tools, PagerDuty, Slack, GitHub, Confluence, and custom APIs.

Market

Relvy competes in the emerging AI-powered incident response, autonomous SRE, observability, and developer-tools market. Y Combinator categorizes it as B2B DevOps AI and describes the product as AI-powered debugging notebooks for incident response. Its closest named competitors include PagerDuty, BigPanda, and Honeycomb, which cover incident management, infrastructure monitoring, and application performance monitoring; more directly autonomous incident-resolution platforms such as Resolve AI are also adjacent competitors.

Relvy is an active, early-stage company from Y Combinator’s Fall 2024 batch. Its public materials claim root-cause analysis for more than 70% of alerts, and its website includes a customer testimonial from Blend. Tracxn reports a $500,000 seed round, while the available public evidence does not establish a dependable revenue figure; the best-supported characterization is therefore an active, funded startup with early customer traction rather than a verified revenue-scale business.