Timeline -
July 2019 - Oct 2021
Role -
UX/UI Designer
With -
1 Project Manager
1 Full-stack Developer
1 ML Engineer
Cool English is an online English learning platform designed specifically for learners aged 9-15, with 1.8 million registered users, 29,000 daily active users, and 360,000 monthly active users (by April, 2024).
After listening the pronunciations and repeating the sentence once, I can usually improve my score to more than 70 points. It really helps me enhance my pronunciation!
Cool English introduces a refreshing approach, it's a fantastic opportunity for children to enhance their speaking skills, conquer fear, and speak confidently and regularly.
I leveraged Google's People + AI Guidebook to design a tailored onboarding experience, ensuring that AI not only added product value but also adhered to user-friendly and ethical standards.
My approach to designing user interactions with the AI-powered English speaking learning platform began with an understanding of the technology's workings and its points of interaction with users.
I integrated AI processes into the user flow, highlighting key touchpoints like data input and output. This created a comprehensive flowchart that combines user actions with system operations, ensuring a seamless, intuitive experience with the AI-powered product.
The redesigned step progress bar shows exact sentences completed and remaining, offering clear feedback and motivating users to finish by visualizing their progress.
The progress bar displayed the user’s completion rate in percentages, giving a general feedback. However, it was too abstract, and users didn’t always feel the significance of their progress and lack motivation.
After confirming feasibility, the two sentences were merged into one, simplifying the interface and making information easier to find and understand.
There were two separate sentences: the first as the practice title and the second generated by the voice-to-text system with AI feedback. This benefited the development but complicated the interface.
Based on teachers feedback, I restructured the design to focus on accurate pronunciation first, aligning features with a step-by-step learning path.
It lacked a clear learning path, providing pronunciation and sentence practice without structured guidance for users to follow.
I rephrased “assessment” to “feedback” and highlighted the AI recommendation section, positioning AI as an assistant rather than an evaluator.
The design featured an “assessment” section where AI provided scores and results. While users valued the insights, the term “assessment” felt too judgmental.