emma Technologies has introduced a set of new capabilities aimed at helping enterprises manage AI infrastructure alongside traditional cloud-native workloads. The update brings GPU compute management, observability, cross-cloud networking, and inference deployment into the company’s existing governance platform.

The launch reflects a growing challenge facing enterprise IT teams: AI infrastructure is rapidly becoming production infrastructure, yet many organizations still manage GPU environments separately from the rest of their cloud operations stack. As AI deployments expand, enterprises are discovering that provisioning GPU capacity is only part of the problem. Governance, networking, utilization, and operational visibility are emerging as equally important concerns.

Enterprises Struggle With AI Infrastructure Complexity

Many organizations now operate AI workloads across multiple cloud providers and infrastructure environments. However, GPU resources are often provisioned manually through provider-specific tooling, creating fragmented operational models and inconsistent governance policies.

This has led to what some in the industry describe as a “Day-2 operations” problem for AI infrastructure. While organizations can acquire GPU capacity, scaling and governing those environments consistently across clouds remains difficult. Challenges around data movement, networking costs, utilization efficiency, and observability are becoming major operational bottlenecks.

“With 76% of organizations already running GPU workloads, making high-performance parallel computing a baseline infrastructure requirement, the need for unified governance frameworks that extend to AI infrastructure, as emma now provides, has never been more urgent,” said Paul Nashawaty. “Our research confirms that despite unprecedented investment in AI infrastructure, organizations continue to encounter bottlenecks related to data movement, orchestration and utilization efficiency, confirming that GPU capacity alone is insufficient for production AI, and that governance platforms like emma are essential to bridging that gap.”

Rather than introducing a separate AI operations platform, emma is integrating GPU and AI management capabilities directly into the infrastructure governance platform enterprises already use for cloud-native operations.

Bringing AI Infrastructure Into Existing Governance Models

The new capabilities are designed as a connected stack that spans provisioning, networking, monitoring, and deployment workflows.

The platform now supports GPU virtual machines and managed Kubernetes clusters under the same governance and policy frameworks used for existing workloads. This approach allows enterprises to apply consistent operational controls across both AI and traditional infrastructure environments.

Observability features provide centralized monitoring for GPU workloads without requiring teams to switch between separate cloud-provider dashboards. The company is also introducing cross-cloud AI networking through its private backbone infrastructure, aiming to reduce complexity and minimize public internet routing costs associated with moving AI workloads between providers.

Another addition focuses on inference deployment workflows. Instead of manually building deployment pipelines for each model, teams can use governed templates designed to standardize and accelerate inference operations.

The broader goal is to reduce the operational fragmentation that often accompanies enterprise AI adoption. Many organizations currently combine standalone GPU management tools, MLOps platforms, and provider-specific services, creating additional operational layers that can complicate governance and increase costs.

AI Infrastructure Is Becoming a Governance Challenge

The announcement highlights a wider shift occurring across the cloud-native computing market. Enterprises are increasingly treating AI workloads as long-term operational infrastructure rather than isolated experimentation environments.

That shift is driving demand for platforms capable of managing AI infrastructure with the same visibility, policy enforcement, and operational consistency applied to traditional enterprise systems.

“GPU infrastructure has been operating outside the governance frameworks that apply to everything else in the enterprise. That’s not sustainable when it’s running production AI. These capabilities bring GPU workloads into the platform that already governs everything else — same policies, same visibility, same operational model. We’re not chasing the AI wave. We’re extending the answer we already had,” said Dmitry Panenkov.

What Comes Next

As enterprises continue expanding AI initiatives, operational concerns are moving beyond model development and into infrastructure management. Organizations are increasingly looking for ways to standardize AI operations across hybrid and multi-cloud environments while controlling costs and maintaining governance.

emma’s latest release reflects a broader industry trend toward consolidating AI infrastructure management into existing cloud operations platforms rather than building entirely separate operational stacks. The success of that strategy may depend on whether enterprises prefer unified governance models over the growing ecosystem of specialized AI infrastructure tools.