> For the complete documentation index, see [llms.txt](https://developer.harness.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.harness.io/roadmap/ai-sre.md).

# AI SRE

Explore planned AI SRE capabilities and enhancements.

This page tracks planned capabilities and enhancements for AI SRE.

### Now

🚧 Q4 FY27 (Nov 2026 - Jan 2027)

<table data-column-title-hidden data-view="cards"><thead><tr><th>Tag<select><option value="incident-management" label="Incident Management" color="red"></option><option value="platform" label="Platform" color="blue"></option></select></th><th>Title</th><th>Description</th></tr></thead><tbody><tr><td><span data-option="incident-management">Incident Management</span></td><td><strong>Problem Tickets</strong></td><td>Promote an incident action item to a Problem ticket in Jira or ServiceNow. Configure Problem Types per project, see every problem on a Problems list page, and view the external ticket's status read-only in AI SRE.</td></tr><tr><td><span data-option="platform">Platform</span></td><td><strong>Audit Trail for AI SRE</strong></td><td>Records who created, edited, or deleted runbooks, schedules, escalation policies, integrations, alert rules, and settings in the Harness Audit Trail. Includes incident and alert lifecycle events.</td></tr></tbody></table>

### Next

Also planned for Q4 FY27 (Nov 2026 - Jan 2027)

<table data-column-title-hidden data-view="cards"><thead><tr><th>Tag<select><option value="platform" label="Platform" color="blue"></option><option value="automation" label="Automation" color="green"></option><option value="ai-intelligence" label="AI &#x26; Intelligence" color="purple"></option><option value="on-call" label="On-Call" color="orange"></option></select></th><th>Title</th><th>Description</th></tr></thead><tbody><tr><td><span data-option="platform">Platform</span></td><td><strong>AI SRE in the New UI</strong></td><td>Moves AI SRE to the new Harness UI design system, with consistent navigation, theming, and dark mode across pages.</td></tr><tr><td><span data-option="automation">Automation</span></td><td><strong>Runbooks on Pipelines (Beta)</strong></td><td>Create and edit pipeline-backed runbooks in Pipeline Studio inside AI SRE, run them from the normal runbook triggers, and read and write incident fields. Available as a beta alongside Classic runbooks.</td></tr><tr><td><span data-option="ai-intelligence">AI &#x26; Intelligence</span></td><td><strong>Shared AI Investigations</strong></td><td>Gives each investigation or incident one persistent agent. Responders in Slack and AI SRE add evidence and ask follow-up questions in the same session instead of starting separate ones.</td></tr><tr><td><span data-option="ai-intelligence">AI &#x26; Intelligence</span></td><td><strong>Investigations from Other Harness Modules</strong></td><td>Starts an AI SRE investigation from a real event in another module, such as a failed deployment or a cost anomaly, without AI SRE setup first. Shows an initial finding, the supporting evidence, and what context is missing.</td></tr><tr><td><span data-option="on-call">On-Call</span></td><td><strong>On-Call Reliability and Contact Readiness</strong></td><td>Improves paging reliability and contact management so the right person is paged without admin workarounds.</td></tr></tbody></table>

### Later

Future quarters

<table data-column-title-hidden data-view="cards"><thead><tr><th>Tag<select><option value="integrations" label="Integrations" color="red"></option><option value="platform" label="Platform" color="blue"></option><option value="ai-intelligence" label="AI &#x26; Intelligence" color="purple"></option></select></th><th>Title</th><th>Description</th></tr></thead><tbody><tr><td><span data-option="integrations">Integrations</span></td><td><strong>MS Teams App</strong></td><td>Native Microsoft Teams application for incident management and on-call workflows.</td></tr><tr><td><span data-option="platform">Platform</span></td><td><strong>Next-Generation Dashboarding in the New UI</strong></td><td>Customizable overview dashboard with widgets for incident metrics, MTTR trends, and alert-to-incident ratio, built for the new UI.</td></tr><tr><td><span data-option="ai-intelligence">AI &#x26; Intelligence</span></td><td><strong>Incident Transcripts as an AI Source</strong></td><td>Makes incident call and meeting transcripts searchable and available to AI tools through MCP, so responders can find what was said during an incident and AI can use it as context.</td></tr></tbody></table>

### Released

What has been released recently

<table data-column-title-hidden data-view="cards"><thead><tr><th>Tag<select><option value="automation" label="Automation" color="green"></option><option value="ai-intelligence" label="AI &#x26; Intelligence" color="purple"></option><option value="incident-management" label="Incident Management" color="red"></option><option value="on-call" label="On-Call" color="orange"></option><option value="platform" label="Platform" color="blue"></option></select></th><th>Title</th><th>Description</th></tr></thead><tbody><tr><td><span data-option="automation">Automation</span></td><td><strong>Create Incident and Create Alert Steps in Pipelines</strong></td><td>Open an AI SRE incident or alert from a pipeline when a deploy, smoke test, or scheduled job fails, with title, severity, service, environment, and execution context filled in from pipeline expressions. Page on-call, name an incident commander, or pass an existing alert ID to update that alert as the pipeline progresses.</td></tr><tr><td><span data-option="ai-intelligence">AI &#x26; Intelligence</span></td><td><strong>AI-Native Postmortems at Incident Close</strong></td><td>Generates an editable postmortem draft from incident fields and timeline activity when the incident closes, and posts it to Slack. Admin-defined templates match your SRE or compliance format, and attaching the review meeting regenerates the draft with its transcript.</td></tr><tr><td><span data-option="incident-management">Incident Management</span></td><td><strong>Action Item Capture During Incidents</strong></td><td>Detects follow-ups in incident conversation, meeting transcription, and notes, and deduplicates them in the incident action list with assignees and due dates when people state them. Say or type "action item" during an incident and a reaction confirms whether it was captured or needs review.</td></tr><tr><td><span data-option="on-call">On-Call</span></td><td><strong>Selective On-Call Migration from PagerDuty, Opsgenie, and xMatters</strong></td><td>Choose which schedules and escalation policies to import and rerun the sync as needed. Harness creates scoped groups, resource groups, and role assignments so team leads can manage their imported on-call setup.</td></tr><tr><td><span data-option="platform">Platform</span></td><td><strong>Resource-Level RBAC for On-Call and Integrations</strong></td><td>Scope access to individual schedules and escalation policies with Harness resource groups, so each team manages its own rotations without account-wide edit access. Inbound webhook integrations are their own RBAC resource type, and users with override permission can set schedule overrides for other rotation members.</td></tr><tr><td><span data-option="on-call">On-Call</span></td><td><strong>Schedule Coverage Digests</strong></td><td>Sends each schedule's owners a daily or weekly email digest of upcoming coverage.</td></tr><tr><td><span data-option="on-call">On-Call</span></td><td><strong>Service Paging and the Page Action</strong></td><td>Give a service its own paging webhook or email address so monitors and jobs can page it, or add a paging action to a runbook. The Page action on On Call and Overview notifies a person before an incident exists (requires On Call), and active pages can be canceled from the responders table.</td></tr><tr><td><span data-option="on-call">On-Call</span></td><td><strong>Mobile Paging Where Google Is Blocked</strong></td><td>The mobile app uses Pushy alongside Firebase to deliver pages and shift reminders in countries where Google services are blocked.</td></tr><tr><td><span data-option="automation">Automation</span></td><td><strong>Nested Conditions and CEL in Rules and Runbooks</strong></td><td>Runbooks, alert rules, transforms, and paging rules support nested AND/OR conditions and CEL expressions. Conditional steps branch on the result, so a staging alert does not wake the production rotation.</td></tr><tr><td><span data-option="automation">Automation</span></td><td><strong>For Each Step in Runbooks</strong></td><td>Runs child steps for every item in a JSON list or template, so one runbook works through every affected host or service without duplicated steps.</td></tr><tr><td><span data-option="automation">Automation</span></td><td><strong>Launch Runbooks from Slack and from Page Events</strong></td><td>Give a runbook a Slack slug and run <code>/harness run</code> with that slug when responders need the procedure. A Page sent event can also start a runbook by channel, escalation policy, or responder, and the runbook can send a configured email update.</td></tr><tr><td><span data-option="automation">Automation</span></td><td><strong>Runbook YAML Editing and Duplication</strong></td><td>Switch the Workflow tab between the visual builder and editable YAML, and duplicate a runbook from the Runbooks list to adapt the copy without changing the original.</td></tr><tr><td><span data-option="ai-intelligence">AI &#x26; Intelligence</span></td><td><strong>Live Root Cause Panel</strong></td><td>Shows AI SRE's evolving analysis in the incident's Root Cause panel or investigation view. When responders state a root cause or rule out a theory in incident messages, AI SRE updates the theory record.</td></tr><tr><td><span data-option="ai-intelligence">AI &#x26; Intelligence</span></td><td><strong>Ask AI from Alerts, Incidents, and Changes</strong></td><td>Open Ask AI from the alerts list or an alert, the incident list or an incident, or the changes list, and the items on screen are supplied as context.</td></tr><tr><td><span data-option="ai-intelligence">AI &#x26; Intelligence</span></td><td><strong>More Signals in Investigations</strong></td><td>The Investigator considers incident-related alerts, recent events from connected change sources, and similar closed incidents. Similar incident search covers incidents closed after the capability was enabled.</td></tr><tr><td><span data-option="ai-intelligence">AI &#x26; Intelligence</span></td><td><strong>Approved Diagnostics for the Investigator</strong></td><td>Admins turn on Expose to AI for selected runbooks in the incident's Harness project. Unselected runbooks stay unavailable to the Investigator.</td></tr><tr><td><span data-option="platform">Platform</span></td><td><strong>SLOs in AI SRE</strong></td><td>SLO pages and settings are generally available in AI SRE as SRM sunsets.</td></tr></tbody></table>

{% @harness-feedback/feedback module="roadmap" pagePath="roadmap/ai-sre" %}
