FIVA
Designing an AI-powered fitness platform that personalizes workouts, nutrition, and recovery to align with each phase of the menstrual cycle.
Project Type
Capstone
​​
​
​
Role​
​
Product Designer
Duration
12 Weeks
​
​​
​
Team​
​
7 GBDA Students
​
​
Project Overview
​​​
Focusing on challenging biases that often arise in today's digital age through technology and design—our Capstone team developed FIVA, an AI-powered fitness and wellness coaching platform tailored for young women and fitness beginners. This 12-week long Capstone aimed to create a more inclusive and supportive digital space, especially for individuals who are often underrepresented or overlooked in fitness technologies.
​
​
What is FIVA​​​
​
FIVA is an AI-powered fitness and wellness coaching platform designed to support young women and fitness beginners by providing personalized health guidance aligned with their menstrual cycles. The platform uses machine learning to generate adaptive workout, nutrition, and recovery plans based on cycle data, health goals, and user behavior over time. FIVA is built with inclusivity in mind—especially for users with irregular cycles or conditions such as PCOS—aiming to bridge the gap in traditional fitness technology that often overlooks the unique needs of women. The system features a clean, user-friendly interface with a personalized onboarding flow, cycle predictions, wellness tracking, and motivational coaching. FIVA promotes an empowering and bias-conscious approach to digital wellness, where technology adapts to the user—not the other way around.​
​
​
​​​
My Role ​
​
-
Led end-to-end user research, including surveys, interviews, and stakeholder mapping
-
Created a comprehensive user journey map to surface pain points and shape product direction
-
Conducted usability testing on high-fidelity prototypes, developing task flows and synthesizing feedback
-
Designed a service blueprint to align user needs with technical and operational feasibility
-
Produced financial projections to support long-term scalability and business value
-
Co-authored a final report summarizing our research, design process, and key insights
​
​
​
Proposal: Personalized Fitness App for Young Women and Fitness Beginners​
​
Project Problem Statement: Young women and fitness beginners often lack personalized exercise plans that align with their hormonal cycles and fitness goals. Most apps offer generic or static recommendations, failing to address their unique needs, resulting in a gap in actionable, science-backed guidance.
​
​
Bias Addressed: The app focuses on bias in fitness and health guidance. Traditional apps often provide generic solutions, failing to account for individual hormonal fluctuations, fitness goals, and the cultural or regional needs of young women starting their fitness journey. Many existing solutions inadvertently reinforce exclusion by not addressing personalization gaps in fitness routines for hormonal health.
​
​
​
Hypothesis: Personalized fitness plans tailored to menstrual cycles and hormonal health will improve young women's fitness adherence, satisfaction, and results.


Solution Statement: The proposed app leverages AI to create individualized fitness and wellness plans, addressing the unique needs of young women starting their fitness journey, with a focus on hormonal health and specific fitness goals.
​​
RESEARCH QUESTION
How can fitness and wellness plans be tailored to hormonal phases to create personalized, goal-oriented fitness experiences for young women and beginners?
Summary of Design Process​​
​
Our design journey began by identifying gender bias in fitness technology; most platforms are built on male-centric models, overlooking the need for not only educational but also personalized, cycle-synced training that adjusts to women’s hormonal fluctuations and menstrual health. During this Identify & Explore phase, we conducted secondary research and the initial system map into the problem, defining FIVA’s direction as a service built to close this gap.

In the Empathise & Discover phase, we conducted surveys and interviews, developed a user persona, and created multiple stakeholder maps to understand FIVA’s role within a larger user journey and service network. MVP and mentor feedback in the Reframe & Define phase revealed that users needed greater clarity on FIVA’s value as a service. Prompted to address this, we created more targeted stakeholder maps and refined our service model to emphasize continuous value delivery. These expanded maps helped define how FIVA operates within a broader service network, supporting users, healthcare providers, and future B2B partners like women’s gyms or wellness clinics. In the Ideate & Develop phase for our HFP, in response to the feedback and user testing, we refined onboarding to clearly communicate value and trust, touchpoints like daily check-in/progress tracking, wearable integration, and added gamified motivation (e.g., streaks, challenge badges) for ongoing engagement and daily value delivery.
Based on feedback on the need for AI, we ultimately chose to keep it as a core feature for real-time personalization. Our user research revealed a strong demand for adaptive tools that responds to daily shifts, which competitor's static apps fail to address. We are committed to training our AI system with diverse datasets and ensuring outputs are medically validated, evolving with users. Each design decision emphasized continuous value delivery and clarified how the service adapts over time to meet women's real, changing needs.
Research Process and Insights​
​
Our research process began with identifying women's needs, specifically personalized fitness plans based on their menstrual cycles, while addressing the gender bias often present in traditional fitness solutions that overlook the unique health of women. We analyzed competitors to understand the lack of integration between cycle tracking and fitness recommendations. User surveys and interviews revealed gaps in the market, revealing that many women feel frustrated that they do not adapt to their changing hormonal levels.
​
Target Audience​
​
Our target audience consisted of women (18-35) looking for customized fitness solutions that align with their menstrual cycles. Key insights include:​
-
Personalization: Users expressed the need for personalized workout plans and nutrition guidance based on their menstrual phase, which traditional apps overlooked.
-
Real-Time Adjustments: Users were looking for an app that could adapt workouts based on their daily fluctuations to adjust the intensity of exercises.
Insights Through User Testing​
​
Onboarding Process:​ Users found that transitioning straight from the splash screen to the login page made them unaware of the app's purpose. We added an onboarding section with informational screens that outline the app’s key features and benefits.
Protein Shake Feature: Users were confused about the protein shake feature, where users did not fully understand the purpose. We refined the design, showcasing the shake deck on user profiles, including completed shakes and in-progress shakes, ensuring easier understanding and tracking.
Apple Watch Integration: Users liked the integration with Apple Watch; however, there was confusion around how users without an Apple Watch could input their information. To address this, we included a preferences survey list to allow users to add their fitness preferences and menstrual details.
Grounded Solution:​
​
By integrating wearable technology and user feedback, we developed a solution that offered cycle-based recommendations, including both work and nutrition plans that cater to each user’s unique menstrual cycle. In addition, we made a more intuitive user interface for easier navigation, with clear, accessible features where cycle information and fitness recommendations are prominently displayed for a more intuitive user experience.
Feasibility & Service Blueprint​
​
FIVA uses a modular monolithic architecture hosted on AWS, allowing rapid MVP iterations and future scalability into microservices . Though deployed as a single unit, it’s divided into modules: cycle syncing, AI recommendations, and nutrition that can be updated independently. This hybrid model balances development speed with the flexibility needed for personalized health services.
One obstacle we anticipated was bias in generalized training data. To address this, FIVA’s AI will be built using TensorFlow or PyTorch and trained on diverse health and fitness datasets to ensure inclusive, medically validated outputs. Transparency features allow users to understand and adjust AI recommendations, building trust. To handle the privacy obstacle, the backend, built with Node.js and hosted on AWS, uses secure, encrypted storage and integrates with trusted APIs like Apple Health and Fitbit. Cloud-based AI processing ensures efficiency without compromising user data. This architecture enables FIVA to deliver personalized, ethical, and scalable support that adapts to real menstrual health needs.
The service blueprint outlines a personalized fitness and wellness platform that adapts to users' menstrual cycles using AI. The service begins when users sign up and input their cycle and fitness data during onboarding. From there, users can access a dashboard that tracks menstrual phases, fitness insights, with phase-based workout and nutrition recommendations. Core touchpoints include the onboarding for values, daily check-ins, real-time cycle and progress tracking, and gamification for motivation. Back-end processes analyze user data and continuously refine recommendations using machine learning, NLP, and reinforcement learning. Support systems manage data encryption, wearable integrations, customer service, and billing. Together, the blueprint illustrates how user actions, system functions, and support processes work in sync to deliver a seamless, personalized fitness experience.
​​
After curating the user journey map and the feasibility and service blueprint, I created a user flow diagram to visualize key interactions and ensure a seamless, intuitive experience aligned with both user needs and technical constraints.
Prototype and Execution​
​
The feedback received from user testing helped us refine the layout, flow, and functionality of the app, leading to the development of our high-fidelity prototype. This iteration process was essential for the app development, specifically for fine-tuning core features, including cycle-based workout recommendations.
The value proposition centers around AI-driven personalization, which adapts workouts and nutrition plans according to the user’s menstrual cycle. For example, during the follicular phase, the app's AI would adjust workout intensity and difficulty based on the user’s needs.
Our design choices were driven by user research and insights gained through feedback and design research. We utilized user feedback to emphasize simplicity and ease of navigation. Our user interface was kept minimalist and clean to ensure that the interface had no visual clutter.
We avoided overwhelming users with excessive information, instead creating an app that was user-friendly and had a streamlined interface. Main features, including cycle tracking and fitness recommendations, were displayed on main screens to ensure easy access and navigation for users. In addition, real-time adjustments and the cycle information were prominently displayed on all pages to align with the app’s goal.
Screens from Low-fidelity design
Screens from Mid-fidelity design
​
​
Onboarding:
Fitness:
Cycle:
Final Product:​
​​​
​
