Problem
With the abundance of AI tools in today’s world, building products has never been easier. But now in such a saturated market, building intuitive products that users love has become more important than ever.
Solution
NAVA is an AI-powered usability testing platform that measures the intuitiveness of any app or website. It captures data (including attention, hesitation, misclicks, navigation paths, eye tracking, etc.), and synthesizes these into clear UX recommendations. Instead of guessing which parts of the app are intuitive and not, or relying on vague user feedback, NAVA pinpoints exactly where and why users struggle.
How It Works
NAVA provides portals for two user groups: product teams and testers. On the product team’s dashboard, they are able to manage/create different projects, and within each project, run different iterations for testing. When they create a new iteration, they get a unique code, which they then send to the testers. The testers enter that unique code on their NAVA portal, start the task, open the app for testing and complete the task.
As the tester completes real tasks, NAVA actively screen records and uses computer vision to record visual signals. In this process, we used Keywords AI to make API calls to OpenAI’s ChatGPT model for behavioral detection. These signals calculate metrics and information (such as which area of the screen is viewed the most, how long it takes a user to complete a task, etc.).
An AI agent analyzes this data, identifies bottlenecks, explains why they occur, and generate concrete recommendations for designers and developers. The app will use Keywords AI for the LLM model that synthesizes data into understandable explanations. Teams are also able to compare metrics across different iterations to quickly see the most intuitive parts of each version.
With the prevalence of prototyping with AI tools, NAVA can easily be added to any product owner’s development process to guide intentional improvements.
Log in or sign up for Devpost to join the conversation.