Cost and Sovereignty

For governments, regulated enterprises, and development organizations, "how should we build our AI capacity?" is rarely a pure cost question. It's a question about who holds the data, who holds the chips, who holds the leverage — and how those choices price out over a multi-year horizon.

Strategy frameworks get you to a defensible decision on the sovereignty-versus-cost tradeoff. But a decision without numbers is a slogan. Once an organization decides how much sovereignty it actually needs, someone has to size the workload, price the compute, and stress-test whether the chosen posture is affordable.

These two tools are built to be used in sequence.

Gen AI Strategy Simulator sits at the strategy layer. Given a starting point, it lets you compare deployment archetypes — full cloud, hybrid, local, edge-plus-central — across supply chain exposure, financing structure, and the adopt/adapt/advance maturity arc. It's built to support the conversation a CIO, finance minister, or board has before picking a vendor or a posture.

Compute Cost Calculator picks up where strategy stops. Once the tradeoff is made, the calculator demands hard numbers: a target population, queries per user per day, model size. It sizes the workload, prices it across deployment patterns, and shows where the sovereignty posture you chose either holds up at scale or collapses under its own cost curve.

No secret sauce. Every assumption is a slider you can move. Both tools are released under MIT — the goal is to make the math contestable, not to win an argument.