A platform operating model organises an enterprise around a set of standing platforms, each responsible for one customer service, product or shared capability. The platform holds its own people, funding, technology assets and applications, keeps them year after year, and is judged on outcomes rather than on delivering a scoped project on time. DBS Bank reorganised into 33 platforms in 2018 on this basis, grouped into business-aligned, enterprise support, shared and enabling platforms.
The hard part is the money and the scorecard. Funding moves from project business cases to an annual platform budget, which frees teams to change things without returning to a committee each time but takes away the finance function’s familiar control points. Leadership has to be joint, or the business and technology sides fall back into client and supplier. And a customer’s problem usually crosses several platforms, which is why DBS later added a journey layer across them.
For AI, the model matters because it gives every use case a standing owner. A model built inside a platform has someone accountable for the business result it is meant to move, which is the precondition for measuring its value at all.