Pause beats a fake win.
A public method for a fictional infrastructure holding group of 50,000 human workers. We show the diary, the gaps, and the refusals. We do not sell a transformation.
Why the GitHub exists Open the diary
The repo is a lab, not a product
The public GitHub is where the method can be inspected. It is not a dump of a finished product, and it is not a claim that adoption has paid for itself.
- Sample data. Synthetic rows you can read and regenerate. Counts are counts. They are not a result.
- Ideas in the open. The diary and the standing issues show what we tried, what we paused, and what is still unknown.
- Methods. Engagement gates, scorecards, an org map, and bridges from O*NET to ESCO. A label is not a benefit.
- Frameworks. Approaches to AI adoption that watch cost, risk, and engagement and enablement together. Pause beats a fake win. Read the methods page.
- BeatAPI lab. Decision, web-research, and optional scoring calls stay off this page. The catalogue is a static snapshot. The key is not in the browser. Open the BeatAPI lab.
Three surfaces, one story
- frontierorg.work carries the method, the diary, and the scorecards. You are on that surface.
- inside.frontierorg.work is a public mock intranet. Same scenario ids. No access control. Marked simulated.
- anorthernman-developer/frontierorg-diary holds the fixture, the schemas, and the generator. Inspect the join before you trust a sentence.
- The org map is a filtered O*NET 31.0 slice with heuristic department tags and, where the official file has a row, an ESCO v1.1.0 link. Open the org map. Labels are not benefits.
- The people page puts three fictional roles in front of that map. Open people. Day 12 records three simulated sessions. Workforce consultations: zero.
The board triad
Customer outcome, employee transition, and risk and cost weigh the same. A hole in one is a pause, not a footnote.
| Weight | Question | Actual |
|---|---|---|
| Equal | Did the customer get the restoration? | not measured |
| Equal | Does a named worker know what changed? | not measured |
| Equal | Can finance and risk defend the cost? | not measured |
What we refuse
- Fifty thousand real directory accounts. The directory is simulated only.
- A claim that this is Cheung Kong Infrastructure, EY, Accenture, Microsoft, SAP, Workday, or Halo.
- An efficiency percentage with no evidence underneath it.