SimuSpeech: an AI-powered platform for public speaking practice

Helping users build confidence through real-time audience simulation and feedback.

Project Overview

Public speaking is one of the most common forms of social anxiety, especially among college students.

This project explored public speaking anxiety through deep-listening interviews and psychological research, then used Kano Model analysis (78 participants) to prioritize UX prototype concepts. The design was evaluated through remote usability testing and SUS surveys (4 participants).

Key Research Questions

- Why do higher-education students experience public speaking anxiety? - Which concepts should be prioritized for development and testing?

Project outcome

My Role

Design Researcher, UX Designer

The Team

Individual Project

industry

AI Tools

Public Speaking Practice Tools

Food & Beverage

Design attributes

Evidence-based Design

Human-AI Interaction Design

Skills

Kano Model Analysis · Remote usability test · System Usability Scale

tools

Figjam, Figma, Google Form, Google Excel, Loop 11, Amazon Mechanical Turk

timeline

Sep - Oct 2024 (7 weeks, part-time)

W 1

W 2

W 3

W 4

W 5

W 6

W 7

The design process

Contextual Research

Contextual Research

Generative Research

Evaluative Research

Evaluative Research

Analysis and Synthesis

Prototype

Ideate

Prototype

01

Contextual Research

Literature review

Psychological framework review

02

Generative Research

In-depth interviews (n=4)

03

Analysis and Synthesis

Thematic analysis

04

Ideate and Prototype

Concept generation

Flow-focused wireframes

05

Evaluative Research

Survey (n=78)

Kano model analysis

Remote usability testing (n=4)

System Usability Scale

design process highlights - evaluative research

Using Kano Model Analysis to prioritize eight conceptual features for prototyping.

Kano-style survey

I conducted a Kano-style survey via Amazon Mechanical Turk (n=78) to evaluate eight conceptual features. Each feature was assessed using functional and dysfunctional questions, and an importance rating.

Kano Model Analysis results

Next, I analyzed valid responses using Kano Model Analysis to inform decisions on which features to prototype and test.

Evaluating the flow prototype through remote task-based usability testing.

This evaluation captured behavioral performance (task completion), perceived usability (SUS), and qualitative feedback (think-aloud) to understand both what happened and why.

Remote task-based usability testing

I conducted remote task-based usability testing on Loop11 to evaluate whether participants could successfully complete task and identify usability issues.

75

Good

System Usability Scale (SUS)

After participants complete their task, they were asked to fill the System Usability Scale (SUS) survey to assess overall perceived usability.

See the work in detail

Thank you for wandering through my little garden of work!

I'm excited to discuss how we can grow something amazing together.

Xuci Hu © 2025

Made with love and lots of digital fertilizer