Overview
Enterprises like banks, insurance companies and hospitals are part of our everyday lives, but we rarely stop to think about what keeps them up and running. Software we use everyday is powered by underlying machines and infrastructure that ensure we are able to use these services without having to think twice.
Teams of people, typically known as system engineers, are responsible for maintaining and running the technology that powers these services. IBM has created a platform and suite of products, known as Cloud Paks, to help enterprise companies with use cases like security, AI, and cloud management.
A key area of focus for Cloud Paks was helping users to troubleshoot issues within services and receive help and support if needed. Many IBM services included ways to do this, but the experiences were disjointed and confusing for users who found themselves using these services together as part of a platform.
I was assigned to lead design in this area of the platform, and build an experience from the ground up. To do this, I asked myself, “How can I help system engineers with varying levels of experience troubleshoot problems in their IBM software?”
The Task
Help site reliability engineers gain a holistic understanding of a problem and determine an appropriate solution as quickly as possible.
Deliverables // Service framework, UX & UI Designs
Category: User Interface, User Experience
Tools: Sketch
Role: Research, Product strategy, UX Design, Visual Design
Research
As the designer in this area of the platform, my first step was to answer a few questions. What is IT infrastructure exactly? How does it all work together? Who’s using this? For what? Where is technology headed? How can I support our users goals moving in that direction?
Some key insights:
Users don’t all have the same amount of knowledge or experience working in IT, and our experience needed to support both extremes.
Daily workflow was riddled with distraction and cognitive overload was a real issue. The services needed to be able to help users understand what needed to be done as quickly as possible.
Just because two people have the same job title doesn’t mean they’re doing the same thing, so experiences needed to be highly personalized to each user.
Keep it simple. Don’t try and reinvent the wheel.
Process
Armed with the knowledge and insights from the initial research, I collaborated with users in product thinking workshops to outline task flows and content that both met needs and fit within the business strategy.
This allowed us to create a framework for troubleshooting and support that could be used ubiquitously across all IBM services; creating a consistent experience across the organization.
User requirements and wishes in hand, I set out to design the first iteration of the product vision. After a few design iterations, I had a prototype that was ready to be tested by users.
Meet Users Where They’re At
First, I created a flow that allowed users to set preferences when it came to alerts and notifications. When an alert was triggered, users had immediate access to critical information they needed to understand the problem and take action as quickly as possible.
What’s Really Going on Here?
Personalized dashboards helped users to understand the root cause of a problem, enabling them to begin to come up with the best solution.
Map it Out
An interactive map visualizing relationships between services, known as a topology, helped users to understand which services were affected, which was critical in deciding which solution might be best to execute.
Help Me Out
An in product help center provided documentation, recommended solutions, and a portal where users could submit support tickets to IBM directly.
MVE
Working with partners in the business and engineering units as well as our users and subject matter experts, I collaborated on a timeline and feature release schedule. I also defined what realistically could be included in our MVP.
Our initial release focused on
helping users troubleshoot common issues on their own
gathering accurate diagnostics to submit with support tickets
User Testing
More rounds of user testing helped us to improve our initial designs. Notable updates included:
Simplifying some design patterns to better serve someone who was distracted
Focusing insights content on metrics that would be most helpful in diagnosing a problem quickly. I also adjusted visuals to help users scan the dashboard more quickly.
Creating designs that met the needs of novice and expert users alike.
Project State
As of January 2020, the first two releases of serviceability features have been implemented in the platform, and the framework I created has been distributed to IBM product teams.
Business results are contingent on user implementation feedback, which is expected to be conducted early in 2021.