AI Fitness Coach

App Design in GoMore (Nov 2020 — Jun 2021)
Project Overview
Focused on professional sports science and AI-powered coaching solutions, this project leverages patented physiological indicators to enhance the intelligence of wearable devices and fitness equipment.

The AI coach is designed to prevent injuries while building personalized fitness models, ensuring tailored and effective training outcomes.
My Contributions
Assisted the business team with client proposals and analysis reports, as well as drafting specification documents and bilingual text strings. Designed information architecture and workflows, collaborated with sports scientists to validate indicator charts, and created hi-fi prototypes for demo day presentations.

Project Participants

Collaboration
・Product team (PDM, PJM, TPM)
・Engineering team (App, Algorithm, QA)
・Sports Scientist team

Stakeholder
・Business team (Sales, MKT, FAE)
・CEO

Project Process

Throughout the process, I transitioned into a product manager role—after designing the UI, I started writing PRDs and attempted to run Scrum development until the product launch.

Project Objectives

Design an AI coaching module that packages the company's "physiological indicators" product, showcasing its value through personalized training insights and performance optimization for clients.

Research Insight

We conducted a survey on both people who have a habit of exercising and those who do not to understand their motivations for exercising as well as the obstacles they face.

Competitive Research

Based on the conclusions drawn from the survey and the advantages of our existing product (algorithm-driven physiological indicators), we analyze whether competing products have a competitive edge in training program design and explore how we can innovate in our approach.

Design Concept

By generating personalized training goals, we create a structured, adaptable, and effective training program for users.
During the training process, an AI coach provides real-time feedback, and after each session, it offers performance-based evaluations.

Before the Training

Before starting a training plan, users need to set parameters and input their physiological indicators through a heart rate monitor. This allows the algorithm to generate a personalized and suitable training plan.

During the Training

Throughout the 4-to-8-month training program, users can train on workout days and fully rest on recovery days.
The AI coach continuously adjusts each session based on the user's performance and physiological data. During workouts, the training difficulty is also dynamically adapted in real-time according to the user's current physical condition.

End of Training

At the end of each training session, a detailed report is generated, analyzing various physiological indicators—these serve as core product features.

Challenges

1. Indicator Design & Application:
Determining which physiological indicators to apply in different scenarios required an in-depth study of the underlying exercise physiology and algorithmic logic.

2. Collaboration with Sports Scientists
Extensive communication was needed to determine the most suitable chart representations for physiological indicators. Additionally, participation in sports experiments was necessary to validate the feasibility and safety of the program.

3. Transition to Product Management
During the early design phase, team departures led to a transition into a PM role. This involved learning to write PRDs after designing UI mockups, leading the development team, and making strategic decisions when facing unrealistic timelines.

My Learning

1. Industry-Specific Knowledge for B2B:
I learned that the B2B industry requires extensive domain knowledge and research to communicate effectively with colleagues and address client pain points accurately.

2. Designing for a Unique Product:
Since the product's core value lies in its algorithm, the design approach differs significantly from B2C products. It requires integrating the client’s vision into the design process.

3. Bridging PM and UI/UX Perspectives:
Transitioning to a PM role helped me understand the differences between a PM and a UI/UX designer's perspectives, particularly in prioritizing development sequences and QA. This cross-functional experience has enriched my skill set.