CapitalOne Design Development Internship

June 2017 - August 2017 | Dallas, TX

In the Summer of 2017, I was fortunate to be a design intern for Capital One, working in their Financial Services line of business. I spearheaded the design side of a project I named Artemis, a tool that assists call center agents troubleshoot customer issues with Capital One's web and mobile platforms. I’m thankful for these learnings and to have worked alongside an amazing design team who showed me the ropes and never hesitated to help when I asked!


I started Artemis off by meeting with the project team that I would be collaborating with on the development and management aspects of this project. The weekly pull-up sessions with the team helped to define the scope, timeline, and overall goals of the project.

The overall goals:

  • Keep customers digital and ensure they didn't need to call in to troubleshoot their issues

  • If customers did need to call in, make sure they can get right back on track

  • Ensure customers have a seamless enrollment experience when opening an account with CapitalOne

My first step was to interview call center agents to determine what their pain points were, what tools enabled them to do their job, and what held them back. I had lovely conversations with over 10 different call center agents, all across different specialties. Some dealt with collections, others with general customer problems, and one team, called "digiCAT," specifically dealt with troubleshooting online issues. 

The interviews were incredibly rich with regards to what they uncovered. Customers were frustrated with the web/mobile experiences and took out their frustrations on the agents, who were trying their hardest to help the clients.


Analysis + Ideation

I took all the insights, quotes, and other interview minutiae and affinity clustered them to better understand the underlying themes. These were the real core issues of what I wanted to tackle with Artemis:

  1. Agents do not have an existing tool for troubleshooting built in to their existing call-center software

    • As a result, they have to resort to 14 (!) external applications to solve a variety of different digital issues

    • The contextual information (like activity notes, browser sessions) are not readily viewable during the call

  2. Agents also need to play hide-and-seek with regards to finding the right tool for each issue, in addition to troubleshooting the issue

  3. The overall troubleshooting flow is confusing and inefficient, causing frustration for both the customer and agent

Despite having such rich information from the interviews, I was challenged by how to show these insights within the context of the agent and customer experiences. I decided to draw the customer journey through their entire CapitalOne end-to-end to better identify what types of issues occur when. Why do customers become "digitally derailed" from the online and mobile experiences? This helped to apply the issues that customers encountered to a specific time point in their journeys.

On the flip side, I wanted to visualize how the agents' experience is when attempting to troubleshoot a customer's issue and how confusing that process is. Both drawings not only helped me to see the end-to-end processes and experiences of agents and customers, but also served as communication tools to my product team and my coworkers.

digiCAT flow.jpg

Before I even started my ideation and prototyping process, I collaborated with my coworkers and learned about a project called Zeus that was set to replace the existing and incredibly outdated call center software. Zeus was designed for the call center agents across multiple different specialties, allowing for flexibility through customizable drag-and-drop widgets that enabled agents complete tasks on-the-fly with the context they needed right in front of them. I decided to use Zeus as the framework that I would be designing within.

The previous software and interface

The previous software and interface

The new software and interface, Zeus

The new software and interface, Zeus


I used the interview insights to determine what tools and information was essential to the call center agents when they were troubleshooting an issue. Beyond that, I also made sure to wholly understand the qualitative needs of the agent experience: an experience that was fast, effective, and easy to understand. I jumped right into determining how I could define this experience within the Zeus framework and decided I would build this tool as a standalone widget. That way, it could be moved around and changed on an as-needed basis. I shied away from using any disruptive elements, such as modals or pop-ups, because they were either ignored or disturbed the agent's flow through the process.

I took some of my low-fidelity sketches and attempted to define the application flow by creating some quick mockups in Balsamiq. I also wanted the Balsamiq mockups so I could conduct some paper prototyping sessions with the call center agents I had interviewed. 

Final Prototype

The final prototype incorporated the core elements of the desired experience I got from my research. I used Invision to make a workable and interactive prototype that could be later tested and refined when added to the roadmap.