Is the new platform
ready?
Moving the report without breaking it
The last page tracks a move. The report is being rebuilt on , and the old Power BI version is the it must match. This page decides whether the new one is ready.
Is 99% good enough?
143 of 144 rows match. That's 99.3%.
Predict first: is that good enough to switch to the new platform?
A grid of differences
| Commodity | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aluminium | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Boron | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Cobalt | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Copper | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Gallium | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Graphite | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Lithium | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Manganese | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Nickel | 0 | 0 | 489,350 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Rare earths | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Tin | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
This is a : each cell is the rebuild minus the reference for one commodity in one month.
Predict first: why show the differences instead of the two sets of values side by side?
Reading a break
Every cell is 0 except one: Nickel, March 2025, 489,350 EUR m. That's nearly 490 billion euros of imports, in one month, for one metal.
Predict first: what kind of cause is most likely?
Two versions at once
The verdict says reference stays live. That's a .
Predict first: why keep the old Power BI version running?
What moves, and what stays
| Item | Where it lives today | Fabric target | Move? | Evidence |
|---|---|---|---|---|
| ECB FX pull | Power Query Web.Contents in the .pbix | Fabric pipeline on schedule, landing in Bronze | Yes | Row count per business day |
| HHI (import-weighted) | DAX measure with SUMX over groups | Keep in semantic model | No | n/a |
| Import reliance % | DAX measure, volume-weighted | Gold fact column, computed in Dataflow Gen2 | Yes | Row-level compare of 12 commodity values |
| Price index TTM change % | DAX time intelligence | Keep in semantic model | No | n/a |
| Refresh status (Failed/Slow/Within target) | Excel column | Silver derived column | Yes | 30 rows, exact match |
| Release gate open | DAX measure over FactReleaseGate | Keep in semantic model, source table from pipeline run log | Partly | 6 checks, exact match |
| Row-level security by tier | Roles in the .pbix | Roles on the Direct Lake model, groups from Entra | Yes | Four role views screenshot-compared |
The lists every item that moves, and the evidence that it still matches. Two items stay where they are.
Predict first: why do the HHI and the price index change stay in the semantic model?
Three layers today
Every row has to match. Show differences so breaks stand out. Keep the reference live until the evidence says switch.
A migration is done when the comparison says so, not when the new platform is built.
Seven pages, one habit
You've read every page of Critical Raw Materials Intelligence the way a business analyst does: the Basis line before the number, the part with the whole, the scope and the period on both sides of a comparison, facts apart from assumptions, and the checks behind it all.