The Next Horizon: Automation Virtual Summit 2026
Evertec’s VM Rightsizing Automation: From Manual Tickets to One Automic Workflow
Session recap · Andrés Flores and Fernando Sousa
The Bobfather
At a Glance:
- How Evertec handled VM rightsizing before automation, with tickets that could take one day or even more
- The Automic Automation workflow, integrated with VCF Operations, that now handles both alarms and catalog requests
- What happens when a check fails: the workflow stops safely, documents the reason and keeps the ticket open
- What Evertec’s Broadcom DX O2 SaaS dashboard showed: 139 automated runs
Session recap · The Next Horizon: Automation Virtual Summit 2026
Autonomous Rightsizing and More: Zero-Touch IT Delivery with VCF and Automic
Andrés Flores, IT Infrastructure Sr. Manager, Evertec,
In conversation with Fernando Sousa, RTO, Broadcom
Evertec’s rightsizing process worked. Monitoring caught the problem, the service catalog took requests and tickets got created. Then everything waited on a person.
At The Next Horizon: Automation Virtual Summit 2026, presented by Automation by Broadcom and Google Cloud on September 24, Fernando Sousa interviewed Andrés Flores about how Evertec changed that. Their session, Autonomous Rightsizing and More: Zero-Touch IT Delivery with VCF and Automic, is a clear look at VM rightsizing automation in a real shop.
Automated runs
139
Rightsizing executions on Evertec’s DX O2 dashboard
Before automation
1+ day
Per ticket, “one day or even more,” per Andrés Flores
Two triggers
63 / 76
Runs started by alarms vs. catalog requests
Meet Evertec
Evertec’s company slide calls it a leading, full-service transaction processing and financial technology company, headquartered in San Juan, Puerto Rico. Fernando pointed to its role in merchant acquiring and payment processing, a presence in 11 countries and customers in more than 26, and described it as providing mission-critical infrastructure. Changes here need to be right.
Rightsizing the Slow Way
Rightsizing means adjusting a VM’s allocated vCPU and memory to match what its workload actually needs. At Evertec, requests arrived two ways: a CPU or memory alarm from monitoring, or an on-demand request through the service catalog. Either way, a person had to pick up the ticket, identify the VM, review the business conditions, check current and recommended values in VMware Cloud Foundation (VCF) Operations, apply or reject the change, and document it.
Depending on workload and availability, Andrés said, that could take one day or even more. Evertec’s slide put it at 1 to several days. The process worked, but it was slow and manual. The main impact was time and consistency: every ticket took human effort, and it was harder to guarantee the same validation criteria every time. Fernando’s summary: too manual and too hard to scale.
VM Rightsizing Automation: One Workflow, Two Front Doors
Evertec now routes both triggers into one Automic Automation workflow integrated with VCF Operations. Alarms start it automatically, and catalog requests start it once they’re approved.
Why Automic? Andrés said Evertec needed a central orchestration point for a multi-step process, and Automic gave the team a controlled way to connect every step, from trigger to documentation, and apply the same logic each time. VCF Operations supplies the current and recommended values and is where approved changes are executed.
The rules come first. Per Evertec’s slide, the VM must belong to Evertec, it can’t be a database server, and the new CPU or memory values must fit policy.
When a Check Fails, the Workflow Stops
Here’s the part worth borrowing. If a check fails, the workflow doesn’t force the change. Fernando said it stops safely, documents the reason and keeps the ticket open. If everything passes, VCF Operations applies the change, and the ticket is updated and closed.
Andrés put it this way: “Automation does not mean removing control. It means applying control consistently.” Fernando called it automation with governance: “Validate first, execute only when appropriate and document every outcome.”
Seeing the Process, Not Just the Ticket
A Broadcom DX O2 SaaS dashboard, integrated with Automic, tracks the runs. Before, Andrés said, visibility was mostly ticket level. Now the team sees how many executions happened, how many were OK or not OK, what conditions are stopping the process, which trends are showing up and where to improve.
Evertec’s dashboard slide showed 139 automated rightsizing executions, 83 of them OK, for a 59.7% success rate. Alarms started 63 of the runs and catalog requests started 76. A Not OK run is one that stopped on a business or technical condition, and the dashboard ranks the causes. The top three were a VM not found in VCF, memory hot-add disabled on the VM, and not enough capacity in ESXi or the cluster.
As Fernando put it: “Automation generates execution, but observability turns execution into insight.” (Want more on why a green status isn’t the whole story? See our recap of Jon Hiett’s session.)
What Changed for the Team
The most visible change, Andrés said, is that the team no longer works every ticket from start to finish, so people can focus on exceptions instead of repetitive manual work. Because the same validation logic runs every time, the process is more predictable and business rules are respected before any resource change.
Evertec’s business value slide listed six benefits: faster execution, consistent validation, reduced manual effort, end-to-end traceability, business observability with DX O2, and continuous improvement. Fernando added that Evertec wanted a solution it could also reuse for AI guardrails.
The closing slide framed all of this as the Automic Automation VCF Advanced Service, with three takeaways: bridge infrastructure and business outcomes, increase the value of your VCF investments, and get enterprise orchestration right on VCF.
One status note. Evertec’s workflow is running today. Separately, in the roadmap session, Rajeev Kumar said Broadcom is optimizing its products for lifecycle management on VCF and creating deep VCF integration for resiliency and disaster recovery. Those are roadmap items without dates; our roadmap recap has the full list.
Andrés’ Advice for Teams Still Doing It by Hand
Asked what to tell a team still rightsizing manually, Andrés said the opportunity goes beyond saving time. The real value is a process you can control, that runs the same way every time and that you can actually see. If monitoring, ticketing and VCF Operations are already in place, the next step is to connect them through automation and observability.
What This Means for You
How many routine requests in your shop sit in a queue until someone’s free? Evertec’s pattern travels well: one workflow for every way a request arrives, your rules checked before anything changes, a clean stop when something’s off, and a record of every run. For more patterns like it, try our recap of four Automic Automation use cases.
You may already own most of the pieces. After 25 years with Automic Automation, RMT would be glad to look at how your requests flow today.
Still rightsizing by hand?
RMT helps Automic teams build workflows like this, starting with a look at how your requests flow today.
More from the summit
The Next Horizon: Automation Virtual Summit 2026, session by session
- Why AI Projects Fail: Sol Rashidi on the Human Advantage
- MCP Server Security: Agentic AI Without the Chaos
- Intelligent Control Plane in Practice: 4 Automic Automation Use Cases
- Evertec’s VM Rightsizing Automation: From Manual Tickets to One Automic Workflow (you’re here)
- The Orchestration Accountability Gap: Inside the SLA Illusion
- From Black Box to Glass Box: A Blueprint for AI Agent Governance
- The Illusion of Green: Automation Observability Beyond the Return Code
- Broadcom Automation Roadmap: What Shipped and What’s Next