Led design of a task-based service catalog for z/OSMF, reducing onboarding complexity and earning multiple design awards (iF + IDA).

Overview

IBM Z is the enterprise mainframe that powers mission-critical operations for the world’s largest banks, insurers, and airlines. Its proprietary operating system, z/OS, supports massive workloads with security, stability, and scale.

To simplify z/OS for newer systems programmers, my team designed the z/OS Management Services Catalog—a task-based web application within the z/OS Management Facility (z/OSMF). Our goal was to reduce complexity and enable users to complete key workflows without deep z/OS expertise.

My Role: UX Designer → Design Lead (Nov 2019)
Team: 2 visual/UX designers, 1 design researcher, 1 content designer
Timeline: April 2019 – September 2021
Launch: Fall 2021

The challenge

Managing a z/OS environment requires deep technical expertise, complex tooling, and institutional knowledge that can take years to acquire. This steep learning curve made it difficult for IBM clients to onboard new systems programmers and scale critical operations.

The Solution

We designed and launched the z/OS Management Services Catalog—a new task-based interface that helps systems programmers manage z/OS environments without relying on green-screen tools or deep institutional knowledge.

The catalog allows new programmers to execute complex tasks through guided, step-by-step services with embedded explanations and contextual help. Experienced programmers can also create custom services that capture and share their expertise.

The tool integrates with existing governance practices, offering controls for approval, publishing, monitoring, and auditing. Based on self-reported data, new users complete tasks more efficiently and with fewer errors.

The Approach

Identifying the Problem

Through early market research and feedback from industry conferences, we uncovered a key challenge: clients were struggling to hire, train, and retain z/OS systems programmers. Experienced professionals needed a way to help newer colleagues perform critical tasks without years of specialized knowledge.

To better understand user needs, we conducted 45 interviews across 7 client organizations, which led to two primary personas:

  • Zach – a senior systems programmer with 30+ years of experience. He manages complex, high-availability environments and mentors new team members, but is stretched thin and unable to transfer his expertise efficiently.

  • Christina – a newer systems programmer responsible for deploying, maintaining, and troubleshooting z/OS systems. She’s eager to learn, but struggles with dense documentation and outdated tools, leaving her dependent on Zach and frustrated by slow career growth.

This knowledge gap between senior and junior programmers became the foundation for our design strategy.

Zach is the persona that represents our experienced systems programmer and Christina is the persona that represents our entry level systems programmer.

Zach is the persona that represents our experienced systems programmer and Christina is the persona that represents our entry level systems programmer.

Story Mapping, Prototyping & Testing

Story Mapping

Our user interviews made it clear: the challenge wasn’t about adding functionality—it was about making the z/OS experience more approachable for less experienced users like Christina.

To achieve this, we reimagined an existing IBM capability—Workflows—and transformed them into guided servicesthat could walk users through complex tasks with built-in support.

We created a detailed end-to-end story map covering the full experience, from initial setup through task completion and service history export. This helped us break the process into actionable scenarios and user stories, which we prioritized for our first release.

We created a story map of the end-to-end process of using services to complete z/OS management tasks so that we could break down the overall process into scenarios and then smaller stories.

We created a story map of the end-to-end process of using services to complete z/OS management tasks so that we could break down the overall process into scenarios and then smaller stories.

Prototyping

We used the IBM Design Language to inform our design choices in terms of type, color, and spacing, and worked to align with the philosophy of guiding our users in every part of our product. We adopted the Carbon Design System which solved easier problems of consistency and standards, enabling us to take on the more complex problems of our users. Using components from the design system we were able to create mockups quickly and put them in front of users for feedback early and often.

An example of an early design for approving a service to be published to the Catalog and the final version that was informed by user feedback.

An example of an early design for approving a service to be published to the Catalog and the final version that was informed by user feedback.

User Testing & Iteration

Validating the Concept

Our initial concept featured a catalog of pre-populated services authored by IBM to help Christina complete common tasks. However, during testing, experienced systems programmers expressed skepticism. They felt that IBM-authored services wouldn’t align with their organization’s specific environments and preferred the ability to create custom solutions themselves.

In response, we shifted our approach: rather than delivering a comprehensive library, we provided a small set of sample services. These acted as templates to demonstrate how the system worked and served as starting points for creating custom, organization-specific services—striking a balance between guidance and flexibility.

Based on user feedback we replaced the pre-populated catalog with a Welcome page to help get started.

Based on user feedback we replaced the pre-populated catalog with a Welcome page to help get started.

Alpha Program & Insights

After a year of prototyping and concept testing, we launched a 14-week Alpha program to validate the product with working code and real users. We engaged 31 participants—some installed the early release locally, while others accessed it through an IBM-hosted server.

Together, we tested six core scenarios related to creating, running, and managing services. Through 38.5 hours of direct engagement and over 764 feedback data points, we uncovered:

  • 93 high-priority improvements to consider for release

  • 43 UX issues and 50 functional defects

  • Key differences in how our two personas—Zach and Christina—interacted with the product

Zach wanted visibility into the underlying commands and JCL to ensure confidence in what was happening behind the scenes. Christina, on the other hand, preferred simplicity and trusted the interface to surface only what she needed when she needed it. These contrasting mental models directly informed how we structured the experience and UI.

 

Prioritizing for Release 1.0

With limited scope for our initial release, our design team partnered closely with engineering and product management to organize and prioritize Alpha feedback. We mapped the data to key user flows to identify the highest-impact areas. The analysis pointed to one clear priority: improving the experience of creating and managing services.

UserExperienceFeedbackMap.png

Managing design quality

As our engineering colleagues implemented the code my team focused on design quality by conducting a QA (quality assurance) of each scenario. We opened issues to track any UX or visual design issues and those issues were integrated into the approval process to close issues once they were resolved. Every detail mattered and my team was diligent about tracking them.

Crafting an end-to-end experience

z/OS Management Services Catalog is an application within another product and the input for each service is an XML file, known as a workflow so it was essential to extend the experience beyond just our product. We worked to integrate our application with the other z/OSMF desktop functions, provide help documentation, and a Content Solution page, to weave together an end-to-end experience.

Positive outcomes

Customers reported that our product would help them reduce errors, save time and reduce complexity. In fact one of our sample services that will ship with the product, Download SMP/E receive order certificate, yielded a 25% reduction in time on task. 

Our customers were overall pleased with our product, and two customers from US Bank were willing to go on the record and say it, providing us with two quote references for pre-GA use.

“z/OS Management Service Catalog is unique in that it will help rookies like me, who come in with very little knowledge about how to execute a batch job, to accelerate our learning curve by providing services that include the required JCL to complete z/OS management tasks.”

— Steve Yang, Early tenure z/OS Systems Programmer, US Bank


“Nice to know that with z/OS Management Services Catalog, IBM is helping us manage our IT infrastructure by modernizing how our early tenure z/OS Systems Programmers perform z/OS management tasks to reduce the learning curve and boost their productivity.”

— Theak Pel, Mid level z/OS Systems Programmer, US Bank

Reflections & Outcomes

Since its release in late 2021, the z/OS Management Services Catalog has been recognized with multiple international design awards, including the iF Design Award and the International Design Award (IDA)—honoring its contribution to modernizing mainframe management.

Although I’ve since moved on from IBM, I’m proud of the long-term impact this work has had. It continues to serve as a case study in balancing legacy systems with modern UX practices and demonstrates how thoughtful design can simplify deeply technical workflows.

Next
Next

MapLab — Democratizing Healthcare Analytics