How Release Intelligence works

Release Intelligence is a workspace-centric hub for a single question: what shipped, in which environment, with what risk, and how we told customers and support.

Each release is a record with title, dates, owner, team, environment (development, staging, or production), status, type (feature, bugfix, and so on), customer impact, QA notes, deployment and rollback context, linked flags, PRs, screenshots, and optional AI-generated briefs.

You add data by creating and editing releases from Releases in the app. Status and environment drive what appears on the dashboard and in promotion-oriented lists.

Typical statuses include Draft, Scheduled, In QA, Monitoring, Released, and Rolled Back. Your team chooses the status that matches your process; the product uses those values to group work on the dashboard.

Rich text is supported on narrative fields so release descriptions and impact notes stay readable when pasted from docs or tickets.

Period filter

The top metrics, charts, and Recent releases list respect the dashboard period (week, month, quarter, or custom range). They only include releases whose release date falls in that window (when you are not using sample mock data).

Upcoming releases

This list is not limited by the period filter. It shows the next few releases whose status is one of:

• Draft — captured but not yet committed to a launch window.
• Scheduled — on the calendar for a planned ship or review.
• In QA — actively under validation.

How to populate it: create releases from Create release (or edit existing ones) and set status to Draft, Scheduled, or In QA. When a release moves to Monitoring, Released, or Rolled Back, it leaves this list by design.

Staging → production

This queue lists releases whose primary environment is Staging and whose status is In QA, Monitoring, Released, or Scheduled — i.e. work that is still anchored on staging before you promote the same release record to production. When you are ready, update the release to Production (and adjust status) and it drops out of this queue.

Releases by environment

Three columns summarize releases in the selected period for Production, Staging, and Development, so you can compare volume at a glance.

Recent activity

Pulls workspace audit events in the selected period (membership, release edits, integrations, and similar actions).

On a release detail page, AI release intelligence can draft executive summaries, customer impact language, rollout risk, QA checklists, Slack and changelog copy, and overlap signals from linked context (release fields, PRs, flags, and so on), when your workspace has the OpenAI integration configured.

The intended workflow is:

1. Generate a draft from current release context.
2. Review tabs (Brief, Rollout, Comms, Signals).
3. Approve when copy is ready for downstream use.
4. Share — for example post to Slack or queue a subscriber digest where Resend and changelog settings allow.

Approved copy helps keep customer communications and internal announcements aligned with the same release record.

Integrations connect external systems to each workspace: for example Jira for keys and sync, GitHub for PR metadata and webhooks, flag providers, LinearB-style delivery snapshots, Slack, Resend for email, and OpenAI for intelligence drafts. Each integration has its own configuration and health signals in Settings → Integrations.

GitHub can be scoped to multiple enabled repositories and a default repo so PR import accepts full URLs,owner/repo#123, or #123 when a default is set.

When a release is public, it can surface on the workspace changelog and embeddable widget. Subscriber digests depend on Resend and approved AI copy where your workflow requires it — see in-app settings for the exact gates.