As enterprises race to move artificial intelligence from experimentation into production, a new challenge of operational complexity is emerging.
While much of the conversation around enterprise AI over the past two years has focused on GPUs, AI factories and infrastructure investment, many organizations are discovering that standing up production-ready environments remains a slow and highly manual process. Deploying new systems often requires weeks of configuration, integration work and coordination across multiple teams before applications can go live.
At Cisco Live last week, Cisco unveiled a new platform designed to address what many enterprises are discovering is the next major AI challenge: turning infrastructure investments into production-ready environments quickly and consistently.
The platform, called Stack Automation by Quali, was developed by Austin-based infrastructure automation company Quali and is being offered exclusively through Cisco. The solution is designed to automate the deployment of full-stack infrastructure environments spanning compute, networking, storage, security, observability, AI tooling and software layers across cloud, on-premises and hybrid environments.
The launch reflects a shift taking place across the enterprise AI market. As organizations move beyond building infrastructure and begin operating AI systems at scale, attention is increasingly turning to how those environments are deployed, governed, secured and managed once they reach production.
Founded in 2011, Quali has built its business around helping enterprises automate and govern complex infrastructure environments. Its Torque platform, which underpins the new Cisco offering, provides organizations with a centralized control plane for provisioning and managing infrastructure across data centers, cloud environments and GPU-based AI systems. In recent years, the company has focused heavily on the challenges enterprises face as AI deployments grow beyond pilot projects and into production environments.
“Enterprise AI infrastructure has become operationally complex at a scale most organizations were never designed to manage manually,” said Lior Koriat, CEO of Quali, in a statement.
“Organizations are now facing a fundamentally different challenge: not simply how to build AI infrastructure, but how to operationalize, govern, and scale AI systems once they enter production. Stack Automation by Quali was built to give enterprises the operational control layer needed to quickly deploy AI infrastructure consistently, securely, and at scale.”
The platform combines Cisco’s validated infrastructure architectures and deployment expertise with technologies from NVIDIA’s AI software ecosystem. It incorporates NVIDIA AI Enterprise software, including NVIDIA NIM microservices, Nemotron open models and agentic AI development capabilities, while supporting Cisco AI PODs as part of Cisco Secure AI Factory with NVIDIA.
The emphasis on automation highlights a growing challenge for enterprise technology teams. While infrastructure investments have accelerated rapidly, deployment processes often remain fragmented, requiring engineers to manually integrate hardware, networking, software, governance controls and security policies before environments are ready for use.
According to Cisco, that deployment burden is becoming a significant obstacle to realizing value from AI investments.
“No enterprise should burn weeks of engineering time manually configuring infrastructure every time they need to stand up a new application,” said Jeremy Foster, SVP & GM, Cisco Compute.
“Stack Automation by Quali was designed to solve that. Whether you’re deploying a traditional workload or standing up a new AI application on a Cisco AI POD, you should be measuring time to outcome in hours not weeks. That’s what this platform offers.”
The announcement comes as enterprises seek ways to industrialize AI deployment. Many organizations have successfully launched AI pilots, but scaling those projects into production often introduces new challenges around governance, compliance, security and consistency.
Those concerns are particularly acute for organizations operating across multiple environments and geographies while navigating requirements around data residency, infrastructure sovereignty and policy enforcement.
To address those challenges, Stack Automation by Quali uses reusable, policy-driven deployment blueprints designed to automate infrastructure provisioning, software integration, runtime governance and AI environment management. Rather than requiring teams to manually assemble and configure infrastructure components for every deployment, the platform aims to make the process repeatable and standardized.
NVIDIA, whose AI software stack is integrated into the offering, views that capability as increasingly important as enterprise AI adoption matures.
“Enterprises are moving AI from pilots into production, and that demands more than great models, it demands a governed, full-stack platform they can run anywhere,” said John Fanelli, VP of Enterprise Software, NVIDIA. “With NVIDIA Nemotron models and NVIDIA AI Enterprise software, which includes NIM microservices and NVIDIA NemoClaw agent blueprints as the foundation, Cisco Stack Automation by Quali gives customers a fast, secure, policy-governed path to deploy NVIDIA-accelerated AI, from a RAG blueprint to Nemotron models, on their own Cisco Secure AI Factory infrastructure. It’s how organizations turn individual deployments into production AI at scale, with confidence.”
For much of the current AI cycle, the focus has been on acquiring infrastructure. Organizations invested heavily in GPUs, AI factories and sovereign AI capabilities. But as those systems move into production, many are discovering that infrastructure ownership alone does not solve the challenge of deploying, governing and scaling AI workloads efficiently.
As a result, the next phase of enterprise AI may be defined less by building infrastructure and more by operating it.
