Crane AI is an AI-powered web tool that help product development teams build mobile software quicker. It uses machine learning and computer vision technology to understand the kind of apps users are trying to build and uses the data to help streamline the development process by providing the code infrastructure and tasks needed to complete the project.
Due to a NDA, I have omitted sensitive information regarding the product and company in this case study.
When I joined the team on April 2018, most of Crane AI was built out. The product was due for launch in 3 months, and stakeholders needed the UX team to test out their existing platform, and uncover any friction points within the product. Since the product was using innovation technology, stakeholders wanted to see if there are any opportunities to leverage the technology behind Crane AI to change and improve the workflow of product development teams.
I worked alongside with a UX Designer (Michelle Lee), our CTO and development team to redefine the platform’s experience. I was responsible for user research, interaction design, wireframing and usability testing. After the research and ideation phase, I was briefly assigned to the computer vision team and came back later to help with the interaction design and wireframes.
To kick-start our first sprint, Michelle and I developed a research plan to test the usability and navigability of the platform, and to understand how designers, developers and product managers operated in a team setting, in hopes of uncovering areas where we can leverage the Crane AI’s technology to help product teams succeed.
We sent out surveys to our network of software developers, designers and product managers, and recruited 9 participants. We screened participants based on their previous experience within a product team setting because we wanted to understand their workflow and the tools they utilized.
We didn’t uncover any opportunities to increase the utility value of Crane AI and we discovered usability issues that were already expected, but we did not expect to uncover a usability issue that was important.
At the time, users interacted with one of the platform’s main feature by communicating with the system’s A.I. through a chat system that utilized N.L.P (Natural Language Processing) technology.
To allow users to execute a function by “speaking/chatting” with the A.I. may seem like a novelty and revolutionary way of interacting with a machine, however, we found the case to be different. When we tested the interaction between the users and the chat system, we found users disliking it or were confused because it was not an interaction that was intuitive.
In addition to discovering the usability issue with the chat function, users were also having difficulty:
Navigating around the platform; the hierarchy of information was unclear
Understanding how to interact with the platform’s following features:
After presenting our findings to the entire product development team, we agreed to restructure the platform’s information architecture, usability issues and interaction design.
Michelle and I led meetings with the development team and CTO to organize all the issues and tasks on Trello and assigned priority points via (Number Game). We made sure the design and development team held daily scrums to make sure we were on track.
We found users to be confused about the product based on the user research prior me joining the team. The marketing team wanted to frame the product as an innovative A.I. tool. This caused users to be confused about the concept and did not understand the technology’s system process that was going on.
To help users understand the concept of our platform, we set 4 principles that drove the direction of the designs and interactions. These principles were set to help create transparency, trust and confidence.
Where Are We Now?
Due to changes in our company’s strategy, we have put the product launch on hold.