One company, three reconciling structures — the management org, the operating brands, and the legal entities behind every transaction.
A group is never one clean tree. Siemens Gamesa Renewable Energy, S.A. is one platform seen three ways — who reports to whom (org), which brand sells the work (brand), and which legal entity books it (entity). They only reconcile through the office, which is why the same acquisition shows up as a brand here, a leader there, and a data-grain gap on the map.
Pick a business unit or a brand to see the offices underneath and how much of the money is office-grain actual vs AI-allocated.
NaN% is office-grain actual; the rest is AI-allocated from area/region postings while integration completes — shown as an estimate, reconciled to the area total.
No mapped office — this entity reports through a shared/area ledger.
CEO → business-unit presidents + corporate functions. Note: the Fire BU is run through its acquired-brand GMs.
Each acquired brand is also a legal entity that books revenue; the colour rail is its business unit, the badge its integration state.
A single job is booked by a legal entity, sold under a brand, owned by a BU leader, and delivered from an office in a region. AI keeps them tied.
Entity resolution maps each legacy entity/brand/office code to one node, so a number can roll up by any lens — by leader, by brand, by BU, or by geography — and still tie to the same total. Where a recent entity still books at region level, the office and brand figures are AI-allocated and flagged, not invented.