TLDR
- PagerDuty launched Paige, an AI agent that detects, investigates and fixes software issues.
- The company renamed its Operations Cloud the PD Reliability Platform.
- Paige works inside Slack and Microsoft Teams and can act with or without human approval.
- Customers who buy Essential, Plus or Ultimate plans before January 15, 2027 get extra AI Actions free for a year.
- PagerDuty shares closed at $16.71 on Monday and traded at $16.42 in pre-market.
PagerDuty, Inc. announced the launch of Paige on Tuesday. Paige is an AI agent built to detect, investigate and fix problems across software systems. The company shared the news in a press release.
Paige is described as a site reliability engineering agent. It can work with human oversight or act on its own in set situations. The company said it is built on more than 16 years of operational data.
What the PD Reliability Platform Includes
PagerDuty also released the PD Reliability Platform. It was formerly called PagerDuty Operations Cloud. The platform combines signal intelligence, incident management and automation.
The platform connects with Slack and Microsoft Teams. Teams can handle incidents in the tools they already use. PagerDuty said it supports more than 750 integrations.
A console shows a company’s incidents, on-call schedules and AI agents in one place. When an incident starts, the platform creates a dedicated channel and a conference bridge. After the issue is fixed, it writes a post-incident review with action items already filled in.
Paige stores information from resolved incidents. It uses that data to help with future responses. PagerDuty said each resolution is written back into shared memory.
How Paige Works During an Incident
Paige acts as a virtual responder in a company’s escalation policy. It can lead a group of supporting agents. These include a Scribe Agent that keeps records, a Shift Agent that runs the schedule and an Insights Agent that offers recommendations.
The agent pulls in alerts and telemetry from across a company’s tools. It also links to Claude and Cursor through MCP for custom setups. It posts a summary for late joiners so calls do not stop.
Paige can suggest fixes based on similar past incidents. It can run runbooks, such as restarting services or running Kubernetes remediation. It can do this across cloud, hybrid and on-premises systems.
Customers set what Paige can see and do for each user group. They decide which actions need human approval. Paige shows its reasoning for each step so teams can check or override it.
Autonomous resolution of well-understood incidents is listed as coming soon. In that mode, Paige would handle incidents the customer defines without waking the on-call team.
PagerDuty also described two experiments in PD Labs. Deep Investigation tests several possible root causes at once and ranks them by confidence. SRE Workbench lets a team and Paige work from the same data on complex incidents. Both are in preview.
PagerDuty also said it established PD Labs as an innovation center. Customers can use it to test features still in development. The company said it invests more than $100 million a year in research and development.
PagerDuty said it manages about one billion incidents per year. It also said about two-thirds of the Fortune 100 use its services. The company cited a finding that 42 percent of organizations reported developer burnout as a direct effect of major incidents.
The platform has a base fee. That fee includes a set number of events, actions and user seats. Customers can buy more capacity through flexible credits.
Organizations that buy Essential, Plus or Ultimate plans before January 15, 2027 will get extra AI Actions capacity at no added cost for the first year.
Paige and the PD Reliability Platform became generally available on Tuesday. PagerDuty shares closed at $16.71 on Monday, down 1.36 percent. In pre-market trading, they were at $16.42, down 1.74 percent.
