Centralizing Product Updates with Automated Pipeline

This release launches the Automated Changelog Pipeline 1.0, transitioning the system from a local proof of concept to a production-ready internal service. The pipeline automatically ingests completed Jira tickets, drafts plain-English summaries using AI, and routes them to a dedicated review board before publishing directly to Webflow. This centralized hub ensures client-facing teams have a single source of truth for all product updates while reducing ad-hoc inquiries to Engineering and Product Team.

Introduction

Internal teams previously spent significant time hunting for product release information across scattered channels like Slack, Jira, and email. This fragmentation often left client-facing teams unprepared on client calls and created constant context-switching for developers answering questions about recent changes.

Key Capabilities

• Automatic, duplicate-safe triggering: The pipeline activates as soon as a feature ticket is marked Done, verifies the root ticket is actually finished before drafting, and ignores repeat webhook events for tickets that are already published or in progress.
• AI generation and editor guardrails: Drafts a plain-English summary of each shipped feature, then automatically screens it for factual claims not backed by the source ticket and for internal engineering artifacts like ticket IDs or field names before a human ever sees it.
• Human-in-the-loop review board: Routes AI-drafted updates to a dedicated Changelog Review board where reviewers can read, edit inline, and approve, approve with edits, or reject each draft.
• Reviewer notification and assignment: Posts to a dedicated Slack channel tagging the source ticket's Lead PM by name, or Marwan Akermouch if none is set, and automatically assigns the review ticket to that PM.
• Webflow CMS publishing: Pushes approved drafts live to the internal hub within seconds of approval, with no manual CMS entry.
• Visual asset passthrough: Carries screenshots and video links from the reviewed Jira ticket through to the published Webflow post for UI-impacting changes.
• Size, platform, and visibility filtering with search: Lets users filter the hub by update size (Boulder, Rock, Sand), platform tags such as Core Platform, Partner Platform, and Everflow Pay, and internal or external visibility, alongside a keyword search across article titles and bodies.

Under the Hood

• Automatic revision with a review-side safety net: If the pipeline's automatic screening can't resolve every flagged issue on its own, it still passes the best available draft to the human reviewer.
• Audience-split Sand batching: Runs two separate weekly digest jobs, one collecting that week's internal_only Sand tickets and one collecting customer_relevant Sand tickets, each producing its own single-voice digest article.