AtumAI: A Principled Framework for Agentic Generation of Datacenter Control-Plane Policies
2026-08-03 • Artificial Intelligence
Artificial IntelligenceDistributed, Parallel, and Cluster ComputingOperating Systems
AI summaryⓘ
The authors developed AtumAI, a system that helps automate creating policies to control datacenters. Normally, designing these policies is slow and complicated, but AtumAI turns a plain language goal into a clear, testable problem and then searches for the best policy using smart algorithms beyond simple language models. This approach makes the process faster, reuses knowledge for new tasks, and explores more possible solutions. They tested AtumAI on tasks like where to place workloads, when to scale resources, and how to manage power, and it performed better than expert-made policies.
datacenter control planeagentic AIpolicy designlarge language models (LLM)formal specificationevolutionary algorithmssurrogate modelsworkload placementresource scalingpower management
Authors
Qiushi Lin, Chaojie Zhang, Íñigo Goiri, Aditya Akella, Ricardo Bianchini, Jovan Stojkovic
Abstract
The efficiency of a datacenter rests on its control plane policies. Designing these policies is increasingly hard: the hardware-software stack grows fast, the design space is vast and interdependent, and prototyping a single policy takes months. Agentic AI promises to automate this search. Off the shelf, however, it falls short on three fronts. It is not formal: with no structured, searchable statement of the problem, the search has little structure to exploit and hard constraints are not guaranteed. It is not transferable: each task is solved from scratch, so nothing learned on one task carries to the next. Finally, it is not systematic: relying on the LLM as the sole source of candidates, it explores a narrow slice of the design space and settles into local optima. We introduce AtumAI, a framework that generates datacenter control-plane policies with agentic AI, making the process formal, transferable, and systematic. From a goal stated in plain language, AtumAI autonomously proposes, tests, and refines candidate policies until one satisfies the request. It does so through two components. The Datacenter Task Compiler automates problem formulation: it compiles the request into a formal, machine-checkable, and searchable specification of the task's objectives, constraints, decision variables, and evaluation methodology. The Evolutionary Design Discovery Loop then searches this specification, expanding the search beyond the LLM itself via a diffusion model, an evolutionary algorithm, and a surrogate model. Together, they reduce onboarding a new task from months of engineering to writing its description. We evaluate AtumAI on three control-plane tasks with distinct problem scopes, design spaces, and trade-offs: workload placement, resource scaling, and power management. Across all tasks, the policies generated by AtumAI consistently outperform expert-engineered baselines.