Bridge AI

AI behavioral therapy remote training application (App)

Overview

Design a professional, clean, and easy-to-use behavioral therapy remote training application (App) to help both therapists and parents train and evaluate Special Education Needs (SEN) children.

Responsibilities

  • Stakeholder Interview
  • Pain Point Analysis
  • Persona Creation
  • Mood Board
  • Wireframes
  • UI Design & Prototyping
  • User Testing

About

Bridge AI, a subsidiary of Bridge Academy Education Centre, is committed to improve Special Education Needs (SEN) students’ learning with technology. By using applied behavior analysis (ABA), behavioral therapists guide SEN children to complete various trainings, and then record and analyze their performance.

The Problem

Bridge AI has trained more than 300 SEN children since 2013. Therefore, it has developed more than 10,000 training contents and accumulated a huge amount of data. They would like to have a professional, comprehensive, and user-friendly interface to show the information and reports.


The Process

The design process of Bridge AI App was shaped by the principles of the design thinking methodology.



Discover

Discovery and Research

Understand the users

Define

Plan and Frame

Structuring content

Develop

Create and Iterate

Designing wireframes and prototypes

Kick-off Meeting with the Client

In our kick-off meeting with the client, we learned that the target audiences are supervisors, therapists and parents who use the App to train the SEN children. We also spent some time to understand their workflow, AI technology and reports.

Discovery Phase

Stakeholder Interviews

We conducted 2 in-depth interviews with our primary users (Supervisor and Therapist) to identify pain points, opportunity, and insights.



0interviewees
0mins per person
0questions asked

Important Takeaways

Supervisor

  • Currently using multiple Excel files
  • Difficult to analyze data
  • Time-consuming to generate report and design sessions

Therapist

  • Easy to input data wrongly in current scoring system
  • Time-consuming to setup sessions
  • Inconvenient communication

Persona

Based on the research we set up two personas.


The personas were important to analyze the behavior and the preferences of the users that used the app.


Personas

Problem Statement


How might we enhance data input accuracy for scoring system?

How might we simplify management and analysis process?

Plan and Frame Phase

Ideate

After brainstorming through a few ideas, we defined the following key features for the app:


Data Visualization Icon

Interactive Data Visualization

  • Quick, clear understanding of the information
  • Analysis at various levels of detail
Dashboard Icon

Personalized report dashboard

  • Overview of student’s report
Scoring Icon

Easy-to-use Scoring System

  • Re-design layout for better data input accuracy
Communication Icon

Communication Tool

  • Instant messaging tool for quick and direct communication

Design System

Design System

Create and Iterate Phase

Mid-fidelity Wireframe

Wireframe were created based on insights from interviews. After creating mid-fidelity wireframes, we converted them into an interactive prototype using Figma software.


We tested the idea with users and fix the problems in the early stage.

Wireframe

Check-in Presentation with Stakeholder

At this point, we presented our progress to the stakeholder. They appreciated the effort but had some feedbacks for us:


  • Personal agenda is not necessary
  • Video review page looks too busy
  • Duplicated or unnecessary information on report dashboard

Final Design

We eliminated unnecessary elements and kept all design elements consistent (according to design system) throughout the app to ensure a clean and well-spaced design.

App Screen

Conclusion

What I have learned from this project?

To design a training App for specified professionals I had to dive deeply into understanding the industry and figure out how to present the information in an easy-understanding way. It takes time but inevitable and valuable.



Creating mid-fidelity prototypes for feedback at early stage was crucial to fixing issues with our approach and determining the future design direction.