Quallie.AI
Quallie.AI is an AI-powered platform which turns qualitative data into insights fast. The product is aimed at research agencies who conduct large research projects. Analysis of qualitative data is typically a time-consuming, manual task involving capture sheets and tagging. Users upload audio and video files to create a research repository in Quallie.AI. They can ask questions to generate findings across the research repository in order to analyse at scale.
Contribution
Design consulting
Team
Domingo Barros, Marco Rovagnati, Marina Jorge
Agency
Pleat Studio
Year
2023
The challenge
The early versions of Quallie.AI lacked design integrity which made it difficult to navigate and discover important features. The information architecture needed to be robust enough to cater for nested concepts and take into account the permissioning of different users.
I worked with the CTO to design the information architecture and navigation which would allow users to move between client accounts and projects efficiently. I also established loose guidelines for colour and component usage, which was further elaborated by their in-house designer.
Information architecture
A research agency will typically have multiple customer accounts, each with multiple projects, and each project could house hundreds of interviews. This results in 4-levels of nested concepts which needed to be visually displayed and easy for users to navigate between (Agency > Customers > Projects > Interviews).
The information architecture also needed to make sense in the context of permissioning. Researchers in an agency typically do not have access to all customer information for privacy reasons. Similarly, a researcher on one project should not necessarily have access to all projects for that customer. It was important to allow users to toggle quickly between customers and projects, rather than navigate forward and backward in a folder-like structure. Each project contains the same tab navigation, which allows a user to toggle between different types of analyses.
A key challenge in this product was representing an enormous amount of information in a simple and efficient manner. The left-hand navigation bar can be minimised, so that the analysis area can be maximised. I introduced accordions, typographic hierarchy and cards to give the pages a clear structure. Icon buttons were added to reduce the visual load on the already dense pages.
I worked with the CTO to outline the onboarding process from sign-up to project setup. This flow captures basic information about the user to identify their ICP. The user is guided to verify their email and either setup a workspace or join an existing one.
