Microsoft is spinning up a new 6,000-person unit dedicated to helping enterprise customers actually implement AI. The team brings together engineers, trainers, managers, and industry specialists to guide companies through the messy reality of AI deployment.
Traditionally, big software companies handed off this kind of low-margin systems integration to outside consultancies. But with large language models, CIOs are drowning in deployment complexity and ballooning costs — and Microsoft wants to be the one to pull them out.
Judson Althoff, Microsoft's commercial business chief, says the company will help customers get a grip on AI spending. That includes swapping pricey proprietary models for cheaper alternatives — a move that could rattle competitors like OpenAI and Google.
Stationing engineers at client offices is becoming the latest arms race among AI players. Palantir popularized the approach; now Salesforce, OpenAI, and AWS are all in the game. Just this Tuesday, Amazon's cloud unit pledged $1 billion to field its own deployment squads.