The Challenge
The team was shipping in a regulated environment where a single change took close to two months to reach done. Merge requests waited about two weeks to land, a shared test cycle took three hours, and AI use was ad hoc rather than built into how the team worked.
The Solution
We co-built an event-driven, AI-assisted release workflow, piloted under a Computer Software Assurance (CSA) framework in parallel with the existing SDLC. Three automated phases are split by two human gates, so sign-offs stay with people. Client engineers touched 79% of all merged code.
Integration layer across Jira, GitLab, TestRail, MasterControl, Google Drive, and Slack
Template engine that turns a controlled document into a validated schema
AI drafting agent producing scope, justification, and release summaries from live data
Client engineers authored 25% of features and reviewed or approved 58% of changes (104 of 132 MRs co-built)
Two firsts for the client: its first CSA pilot, and its first AI-generated compliance documents submitted for auditor review in a CLIA environment. The workflow now runs as a shared platform owned by bioinformatics, and a new team can onboard in about a week with no rebuild.
ROI Achieved
13 Release steps automated ~45 hrs/yr recaptured per team; ~2,250 hrs/yr at 50 teams (client projection)
Lower cost per release makes a faster patch cadence viable
Liatrio-computed, based on: 135 min saved/release; 45 h/yr ⇒ ~20 releases/yr
