self-service analytics tools

The hybrid AI control plane is already here, and it runs on three platforms

Enterprises don't choose one orchestration platform, they run three, on average, and they pick them for flexibility, not loyalty to any single model.

3 min readVentureBeat
The hybrid AI control plane is already here, and it runs on three platforms

**Our Take: The Orchestration Generation Gap**

The enterprise has spoken, and its message is clear: the era of the single, all-encompassing platform is officially over. Our latest Pulse Research reveals a sophisticated, deliberate shift toward orchestration as a portfolio strategy, three platforms on average, chosen not for loyalty to a single model, but for the freedom to move between them. This is the mark of a maturing market. The buying logic has inverted from vendor affinity to vendor optionality, with flexibility across models and tools outranking model gravity by nearly three to one. Enterprises are no longer asking which platform will give them the smartest agent; they are asking which one will let them change their mind tomorrow without rebuilding their entire stack.

Yet, within this strategic maturity lies a critical, and costly, blind spot. While the architecture advances, embracing hybrid control planes and investing heavily in monitoring and permissions, the fiscal guardrails remain worryingly primitive. The fact that one in five enterprises admits to having no real-time mechanism to halt a runaway agent before a bill arrives is not just an operational gap; it is a fundamental failure of governance. It is a stark juxtaposition: we are building sophisticated multi-step reasoning engines, yet we are still trying to meter them with post-hoc log analysis. You are investing in the intelligence, but you are leaving the accountability to chance.

This tension, the desire for powerful, autonomous systems versus the fear of their unconstrained cost, is the defining challenge of this generation of AI. The data suggests we are spending to see and govern agents, but the weakest link in the chain is the very system we are trying to control. The value-for-money satisfaction score sits at the bottom of the pile, a silent indictment of a tooling ecosystem that has prioritized capability over cost containment. The platforms have become incredibly good at making agents think; they are just beginning to grapple with how to make them affordable and auditable.

The path forward is not to retreat from the multi-platform, hybrid future. That ship has sailed, and rightfully so. Instead, the next phase of innovation must focus on the control plane itself, not just orchestrating the tasks, but orchestrating the spend. The enterprises that will lead the next wave are not necessarily those with the most powerful models, but those that can build a financial and security architecture as robust as their execution layer. The question is no longer what your agents can do, but what they are allowed to cost, and who has the power to pull the plug before the meter runs out.

From VentureBeat

Across 107 enterprises, agentic orchestration is not a choice of a single platform.

The typical enterprise runs three orchestration platforms at once, and selects them for flexibility across models rather than affinity to any single one. Microsoft leads primary usage while Anthropic leads forward consideration by a wide margin.

Read the original at VentureBeat