Functionunverified
AI and Compute Finance
An emerging function covering the economics of training and serving models — accelerator capacity, cost per token, commitment coverage and the depreciation assumptions underneath them.
Relationships pointing at this node.
owns
15The business-partner org unit accountable for the task.
- AI initiative ROI and payback measurementTask
- AI product gross margin analysisTask
- AI vendor spend rationalizationTask
- Buy vs rent compute analysisTask
- Cloud commitment and reservation strategyTask
- Compute capacity planningTask
- Consumption and token-based pricing designTask
- Credit and consumption revenue forecastingTask
- Datacenter capex and depreciation modelingTask
- GPU cluster unit economicsTask
- Inference cost per token / per requestTask
- Internal compute allocation and chargebackTask
- Model lifecycle cost accountingTask
- Power, cooling and site cost modelingTask
- Training run cost modelingTask
staffed by
1The org unit the role sits in.