Imagine a financial services team getting ready to put a new AI application into production. The architecture is approved. The compute, networking, and storage have been selected. Security requirements are defined. On paper, the hard decisions are made.
But the infrastructure is still not ready to run the application.
Before that happens, the team may have to configure multiple tools, execute anywhere from hundreds to thousands of infrastructure actions, and work through 100 or more handoffs and approvals across different teams.
Each one of those steps introduces another opportunity for delay or error. A configuration doesn’t match the design. One team finishes its work but waits on another. A failed handoff sends part of the process in reverse.
Stack Automation by Quali, co-developed with Cisco, closes the gap between purchase and production by automating the planning, ordering, and deployment of multivendor, full-stack solutions from a single console.
Here’s the question that matters most: How quickly can you turn the infrastructure you bought into a running workload? The full-stack challenge isn’t just what you deploy. It’s how you deploy it.
That financial services example isn’t unique. The same challenge shows up anywhere teams are trying to stand up modern application environments.
You may be configuring servers, networking, and storage. Layering in Kubernetes or other software. Applying security and observability. Integrating technology from multiple vendors. And making sure everything is configured according to the architecture you originally designed.
For an AI environment, that complexity only increases. And this is where an important distinction comes into play: Validation gives you confidence that the architecture works. It doesn’t automatically make it easy to deploy.
Too often, the last mile between “we bought it” and “it’s running” still involves manual processes, specialized expertise, and a lot of stitching things together.
That’s the gap we are closing.
Cisco and third-party software were already available in Stack Automation. Now, you can order full-stack solutions directly from Cisco. This includes blueprints for Splunk POD, FlexPod, and Cisco AI PODs, along with blueprints for the additional software layers to deploy an end-to-end Cisco Secure AI Factory with NVIDIA. The result is a new way to automate the configuration and deployment of Cisco full-stack solutions. With Stack Automation, we’re taking the expertise embedded in Cisco Validated Designs and turning it into repeatable automation across compute, networking, storage, and software. Rather than being the end of the architecture process, the validated design becomes the starting point for an automated, repeatable deployment.
Plan, order, and install a full-stack solution faster #
From the start, Stack Automation can help teams model a solution, visualize the components, and generate the bill of materials before anything is ordered. Once the order is placed through Cisco Commerce Workspace, the order information and components flow into Stack Automation to become the basis for automated deployment workflows. When the equipment arrives, even the physical installation can become simpler. An optional augmented-reality experience can show the installer, right at the rack, where equipment belongs and how it should be cabled, reducing another source of error and delay.
Then, instead of manually building every layer from scratch, teams can start with automation blueprints based on Cisco Validated Designs.
And this is where the meaning of “full stack” matters. Modern infrastructure rarely comes from one vendor at every layer. A FlexPod environment, for example, brings together Cisco compute and networking, NetApp storage, and Red Hat OpenShift. Stack Automation can use an out-of-the-box blueprint to automate deployment of that validated configuration.
The same approach can be applied to Cisco AI PODs for accelerated AI workloads. There, the automation can build from the core AI POD and add software layers, such as Splunk, Isovalent, and Cisco AI Defense, to create a complete Cisco Secure AI Factory with NVIDIA.
Full stack doesn’t mean every component has to come from one company. It means the technologies customers choose can be validated and deployed as one solution.
Shorten the distance from experimentation to production #
There’s another piece of this that I think is especially important: innovation often starts somewhere very different from where it ultimately has to run.
A developer may start in a sandbox, experiment with a configuration, or build an infrastructure-as-code workflow. But moving that work into a real production environment can require an entirely new process.
With Cisco DevNet now integrated into Stack Automation, developers have a more direct path from exploring automation in a sandbox to applying it against real infrastructure. And with agentic capabilities, that path can extend beyond developers and automation experts. Teams can describe what they want to automate in plain language, while Stack Automation helps translate their domain knowledge into automated tasks and solutions.
The goal shouldn’t be to make experimentation easier while leaving production difficult. Every additional handoff between design, testing, and deployment adds time before the infrastructure starts doing the job it was purchased to do.
Simplify operations beyond automation #
This is also why Stack Automation fits into the broader direction of Cisco Cloud Control. We’ve been talking a lot at Cisco about bringing infrastructure together across domains, but a full stack isn’t especially valuable if it simply gives customers more things to manage.
The value comes when the stack behaves like a system.
Stack Automation brings day-0 planning and day-1 deployment automation into that broader experience. Its job is to get infrastructure production-ready and then hand it off to the management platforms customers already use for day-2 operations, including Cisco Intersight and Cisco Nexus Dashboard. Instead of replacing those tools, Stack Automation helps ensure the infrastructure reaches them configured, validated, and ready to operate.
For customers, the vision is straightforward: One experience. A validated architecture. Automation from deployment through operations. And repeatable shouldn’t mean rigid.
Teams can customize the out-of-the-box, validated blueprints from Cisco or create new ones from scratch. In either case, they can attach their own infrastructure-as-code repositories with workflows they already trust, or they can use the built-in AI assistant or agentic tools like OpenAI Codex or Claude Cowork to draw on all available resources and build entirely new blueprints. This gives customers flexibility without forcing every environment into the same mold.
Accelerate time to value #
For a long time, infrastructure conversations have centered on specifications: CPUs, bandwidth, GPUs, and storage. Those things still matter. But increasingly, another question belongs right beside them: How quickly can I turn all of that infrastructure into something useful?
A server sitting in a rack isn’t delivering value because it arrived quickly. A validated architecture isn’t delivering value because the diagram is complete.
The value starts when applications are running, AI models are serving users, and infrastructure is doing the job the business bought it to do.
Customers shouldn’t have to solve the same deployment challenges from scratch every time they stand up a new environment. They should be able to start with designs, configurations, and engineering expertise backed by years of Cisco and third-party experience. Stack Automation makes that expertise repeatable through validated blueprints, helping simplify the path from design to production.
Because, ultimately, customers aren’t buying infrastructure for the sake of infrastructure. They’re buying it to run applications, deploy AI, and solve real business problems. The faster we can help them get there, the sooner that infrastructure starts delivering value.
Realize value faster with cloud-like speed
Explore Stack Automation by Quali