project-proposal-2025

MyFitPlan - A Science Driven Fitness and Nutrition Planner

The fitness industry is plagued with misinformation, making it difficult for those new to the sphere to find and follow some evidence based training and nutrition plans. On top of the misinformation, scientific research in the field is constantly evoling with most fitness apps failing to update their algorithims accordingly. MyFitPlan is an AI powered platform which dynamically generates personalised workout and meal plans based on the latest published research. Through integration with scientific databses, wearable devices and other fitness tracking applications, MyFitPlan ensures users receive accurate, adaptable and research backed guidance.

Author

Name: Benjamin Ghahramani Student number: 47492299

Functionality

Features

• AI Powered Workout and Nutrition Plans - Recomendations are generated and updated in accordance with recent scientific studies.
• Live Science Integration - Pulls data from PubMed, Google Scholar and other research sources to model fitness plans.
• Wearable Tech and Fitness App Integration - Can sync data with apple health, google fit and other devices.
• Myth Busting Chatbot - Answers fitness and nutrition related questions with valid sources.
• Progress Tracking and Adaptive Adjustments - Algorithim will mend plans where necessary based on individual progress.
• Dietary Preferences and Budget Considerations - Customises meal plans based on dietary needs and financial or other constraints.

User Flow

1. Sign up and Onboarding - Users provide fitness goals, current activity levels and dietary preferences.
2. Dashboard - Users can access personalised fitness and nutrition plans, progress tracking and updates based on new research.
3. Workouts and Meal Plans - AI generates evidence based plans, dynamically adjusting with the rapidly growing research.
4. Chatbot - Users can interact with the AI assistant chatbot to answer health and fitness related questions.
5. Tracking and Adjustments - Users would feed the system with their ongoing progress, or synced through wearables, then system would adapt plans accordingly.

Scope

Minimum Viable Product

• AI powered workout and meal planning based on user input of their goals, weight activity level, etc.
• Integration with fitness tracking apps and devices, Apple Health, Apple Watch, etc.
• A scientific research databse with published journals that update the plans appropriately.
• A basic chatbot for answering fitness related questions.
• User progress tracking with simple data visualisation.
• A secure and scalable backend infrastructure to support user activity.

Quality Attributes

Reliability

Reliability is crucial for ensuring that users receive consistent, accurate and trustworthy recommendations. Since MyFitPlan is built on scientific research, it is important that the system maintains a high degree of correctness in both data ingestion and recommendation output. To achieve this, the platform will validate AI generated plans against expert reviewed fitness programs, conduct automated checks on data accuracy from its sources, and use redundant storage to prevent data loss. The reliability of the system will be assessed by monitoring uptime with a target of 99.9%, comparing AI-generated plans with professional fitness programs, and checking chatbot answers against verified sources for correctness.

Interoperability

Interoperability is essential for MyFitPlan to integrate effectively with a variety of external data sources and fitness ecosystems. Since the platform relies on input from wearable fitness devices and external research databases, it must be able to seamlessly exchange data with these systems. This includes standardised data exchange formats, consistent API behaviour, and robust input validation to ensure integrity and consistency. Interoperability will be evaluated based on the system’s ability to correctly retrieve and process information from fitness APIs and research databases, as well as its compatibility with commonly used health and fitness platforms.

Scalability

Scalability ensures that MyFitPlan can support a growing user base without performance degradation. To achieve this the platform will leverage cloud infrastructure with auto scaling, a distributed database and load balancing to manage traffic peaks efficiently. The system will be stress-tested under increasing user loads to determine its maximum throughput and responsiveness. Metrics such as system response time during simulated peak usage, the efficiency of load distribution, and resource utilisation under load will be used to evaluate scalability.

Evaluation

Reliability

To evaluate reliability, the system will be tested under both normal and high-load conditions, ensuring consistent plan generation and accurate chatbot responses. Expert-reviewed fitness programs will be used as a benchmark, and the system’s uptime will be monitored to ensure it meets the availability target.

Interoperability

Interoperability will be assessed by integrating MyFitPlan with multiple external APIs and verifying successful data exchange. This includes retrieving user metrics from wearable devices and incorporating up-to-date research from academic sources. The correctness, latency, and fault tolerance of these integrations will be key indicators of interoperability success.

Scalability

Scalability will be evaluated through systematic stress testing using increasing levels of simulated users. Key performance indicators such as response time, error rates, and auto-scaling behaviour will be monitored and analysed to ensure the system can grow efficiently with user demand.

By ensuring accurate, science-driven recommendations, seamless system integrations, and support for growth, MyFitPlan aims to become a trusted fitness companion for individuals pursuing evidence-based personal development.