The problem
In 2016, Alaska Airlines was in the process of moving off its legacy systems and onto new technology that would free Customer Service Agents (CSAs) from the counter and let them interact with customers more easily. Alaska had chosen the iPad mini as the device to help agents and customers interact in a more fluid, personal way.
The constraint. The counter is not just a tool — it's a psychological anchor for agents under pressure. Removing it required making the iPad feel more reliable, not just more portable.
Approach
At the time, CSAs at Alaska Airlines were using software dating back 20 years to help customers change flights, check bags, move seats, board aircraft, and handle any number of special circumstances and requests. Rather than recreate that existing software and install it on iPads, we decided to take the time to understand what CSAs actually did and how they used their current systems. That way, the design and product teams could focus on building software around CSAs and their needs, instead of perpetuating workflows that were often outdated and inefficient.
Research & sketches
Alaska had already found success building single-purpose applications that helped CSAs complete specific tasks, like scanning bags. The focus of our application was to help CSAs board passengers onto the aircraft. The legacy system for this process was known as the "on-board manager" (OBM), and our goal was to develop an application that would replace it and make boarding easier and more efficient. I spent several days observing CSAs as they boarded planes and conducted numerous contextual interviews to learn what mattered most in helping them complete their tasks.

The counter is not the product. The boarding experience is the product.
Design
Working on paper was a great way to get ideas out quickly so their potential could be evaluated. Once some ideas started to come together on paper, I moved them into the computer and developed a task flow that captured the boarding process and all of the tools a CSA would need to complete their work. Creating wireframes that communicated my initial concepts clearly was the next step. As a team, we had enough information to start assembling the tools boarding agents needed most, and we were also able to spot inefficiencies in the current process that a mobile boarding app could easily simplify.
One of the needs CSAs identified during boarding was the ability to quickly let passengers change seats. CSAs received this request often enough to warrant a mobile solution, especially in smaller airports where there might not be extra resources to help with boarding. A tool within the boarding app that let a CSA change passenger seating quickly and easily was a necessary component of the proposed Boarding App.
Prototype
After the main components of the Boarding App were identified and wireframed, I turned to how CSAs would actually access these tools in the context of their duties. I made the primary task — boarding the aircraft — the focal point of the interaction, while the tasks that "presented themselves" along the way were literally pushed to the side. The result was an interaction that let users quickly switch between tools when needed.
More iterations & not-ons
Another of the boarding agent's duties was verifying that passengers were literally "not on" the aircraft. If passengers failed to show up for a flight by a specific time, they were considered "no-shows" and their seats were released so stand-by passengers could take them. Before the seats could be released, the boarding agent had to physically print out the passenger list and walk it down the jet-way to confirm those passengers weren't on board. This was called the "not-on" list, and the process was tedious and time-consuming for CSAs. Creating a digital "not-on" list that let agents board passengers directly from the plane would be a tremendous time saver, and it was a critical component of the new Boarding App. I created a thin-slice MVP for this specific purpose.
The "not-on" list workflow was the highest-value discovery of the project. Agents were walking back to the counter multiple times per boarding to update standby status. A digital list eliminated that entirely — and agents could complete the check while on the aircraft.