Show HN: Intelix – AI agent that investigates and resolves production incidents

Intelix, an AI agent built to investigate and resolve production incidents without waiting for an engineer to intervene, launched publicly on August 19, 2026, via a Show HN post by its creator, HN user ved_p05.

The tool connects to PagerDuty, Datadog, GitHub, and Kubernetes to gather evidence when an incident fires, according to the Show HN post. From that evidence it identifies and ranks possible root causes, validates whether a proposed recovery action is safe to execute, and then carries out the recovery — all without requiring a human to page through dashboards. The creator also built an interactive live demo on intelix.dev where visitors can watch the agent work through a simulated incident in real time.

Per the Show HN post, Intelix is described as an early version currently focused on automated incident investigation and recovery. No funding, team size, founding date, or pricing details were disclosed in the source material.

The product sits in a growing category of AI-powered site-reliability and DevOps tooling. Engineering teams at larger organizations typically rely on on-call rotations and runbooks to triage production failures, a process that can take minutes to hours depending on the complexity of the system and the availability of the on-call engineer. Automating that first-response loop — pulling signals from alerting, observability, version control, and orchestration layers simultaneously — is the core bet Intelix is making.

The creator notes this is an early release, suggesting the current scope is limited to investigation and recovery workflows, with no mention yet of broader incident-management features such as postmortem generation, stakeholder communication, or integrations beyond the four listed platforms. What comes next was not specified in the source material beyond the framing of the current version as a starting point.