The Billion-Dollar Confession
AWS spent a billion dollars putting engineers inside customer teams.
AWS announced a billion-dollar Forward Deployed Engineering organization yesterday. Thousands of AI engineers embedded directly inside customer teams to co-build and run production systems.
The NFL shipped Fantasy AI in weeks with an FDE team. Southwest Airlines already has one. Cox Automotive, the NBA, the Allen Institute, Ricoh. These are live systems at organizations where downtime is not an option.
Here is what makes FDE different from consulting. AWS does not need billable hours. Every customer running production AI generates platform revenue indefinitely. So the incentive is to get you operational and leave. The deliverable is a working system, a knowledge graph in your own AWS account, and internal staff trained to run it without AWS in the room.
Three companies built the most capable AI models in history. All three arrived at the same conclusion within 90 days of each other. OpenAI started a consulting company. Anthropic embedded engineers with enterprise customers. AWS committed a billion dollars. That investment went toward making the humans around the AI more capable of using it.
Those investments point to a shared constraint: operational capacity. If your organization is still evaluating which model to use, you are solving a problem that was already solved.