Research

Open questions, tested in the open.

Two tracks. Each page says what the question is, how far along it is, and how to take part. These are early and exploratory: no product, no results yet. Updates get added with a date as the evidence comes in.

Early pilot · cases not frozen

AI Reviewer pressure-test

When a person reviews AI-assisted work in a regulated setting, do they catch the errors that matter?

Three synthetic pharma-quality cases. The reviewer accepts, corrects, rejects or escalates each AI output and points to the evidence. Scored on what they catch, what they flag wrongly, and how long it takes.

Last update: 6 October 2026

Question, stage and how to take part
Practitioner conversations · no findings yet

Release evidence and review consolidation

How much manual work still goes into assembling release, validation and traceability evidence after the tools have done their part?

Three lanes: reconciling Jira and test-management exports into a traceability matrix without inventing links, what reconciliation remains in Azure DevOps Test Plans, and final release-evidence assembly across industries.

Last update: 6 October 2026

Question, stage and how to take part