Report 1. Foundation & Planning #

UrTraining #

Authors: Ildar Rakiev, Makar Dyachenko, Salavat Faizullin, Egor Chernobrovkin, Alexandra Starikova, Ilona Dziurava, Anisya Kochetkova

Code repository: https://github.com/IU-Capstone-Project-2025/UrTraining

Date: June 2025

Team Members #

💡 Project Idea #

Problem Statement #

Fitness and health are booming industries, but people are tired of one-size-fits-all solutions. They are looking for personalized guidance that adapts to their needs. At the same time, trainers want to share their knowledge but struggle to package it effectively or reach an audience online.

Moreover, the Internet is overflowing with conflicting advice from self-proclaimed experts, leaving people overwhelmed and confused. This overload of opinions creates paralysis, making it harder for individuals to take action or stick with a plan.

Meanwhile, trainers have valuable insights but lack the digital tools and strategies to grow their reputation and attract clients. Many great ideas get lost, helping no one. Our mission is to cut through the noise, make first steps easier for users, and help trainers connect with the people who need them most.

Competitive Analysis #

Competitors either provide pre-made workouts (with no real coach interaction), serve as promotional platforms (like Instagram* or YouTube), or simply function as marketplaces.

Problem: UGC platforms (like YouTube, Udemy, marketplaces) often face content quality issues. Without filters, trainers might upload irrelevant or even harmful content.
Validation: Most major platforms use content filters — either automated or manual. A basic NLP scanner (even without ML) already helps a lot. Users expect to pay for actual workout programs, not junk.

Problem: Fitness apps like Fitify, Nike Training Club, or MyFitnessPal often lack real customization — users manually look for plans that match their goals and level.
Validation: Personalization is key to retention. Simplified algorithms for recommending suitable workout plans are already used successfully in health/fitness apps. We can apply a lightweight version based on user’s questionnaire data, even at MVP stage.

Goal #

The goal of our project is to simplify trainer-side workout plan uploads while automating user-side matching of optimal training programs.

Solution #

Our solution simplifies workout plan creation for trainers while using AI to match users with perfectly tailored programs. We combine live coaches, smart matching, fast knowledge integration, AI-powered tools, and - most importantly - effortless monetization. Trainers don’t need marketing skills to earn, and clients don’t need expertise to find value - our service handles the selection for them, ensuring they only get what works.

Ideas during brainstorming #

1. Quality check for uploaded workout plans #

Use a lightweight NLP model (or manual moderation at first) to filter out irrelevant content. The workout plan should contain structured training info like sets, timing, rest periods, and proper exercise names.

2. Workout cards instead of long texts #

Structure the training plans into short, clean cards showing key elements: exercise name, sets, rest time, etc. Easy to read. In the future — we’ll add visuals like icons, images, and videos.

3. Basic progress tracker #

Allow users to mark workouts as “done” and maybe leave a note. No calendar integration for MVP — keep UX simple.

4. Trainer verification & trust system #

Trainers can upload certificates. Show “verified” badge on profile. Ratings and reviews will boost trust. Internal badges or “recommended by platform” can highlight new or quality trainers.

5. Workout recommendation algorithm based on user data #

Take into account the user’s goals, level, and any contraindications. Show programs that match their needs. Possibly with a short explanation powered by AI.

6. Flexible program upload for trainers #

Trainers can either upload PDFs, scan paper notes, or fill out structured forms. We’ll process and turn it into proper workout cards.

7. Built-in injury disclaimer and health filters #

Add a legal disclaimer like “Use at your own risk.” Filter programs that might be unsafe for specific user conditions using LLM or basic keyword checks.

Metrics of Success #

To measure the success of our project, we would use the following metrics:

• Feature Completion Rate – Aim to implement 100% of the core features (MVP) defined in our project scope.
• System Stability – During testing, the system should run with zero critical bugs and less than 3 minor issues.
• User Feedback Score – From peer reviewers or instructors, we aim for an average satisfaction rating of 8 out of 10.
• Task Success Rate – At least 90% of test users should be able to complete key tasks without assistance (e.g., following a plan, navigating the UI).

Given the academic nature of the project, real-world adoption is out of scope for now, but these metrics allow us to evaluate the effectiveness and readiness of the system.

👥 Users #

Target Audience #

User Persona #

Trainer Persona – Alex, 32 #

Alex is a certified personal trainer with over 8 years of experience. He currently works freelance and has built a small following on social media, where he shares workout tips and motivational content. Despite his expertise, Alex struggles to scale his impact online. He lacks the time and technical skills to build structured digital programs or run marketing campaigns.

Goals:

• Share his knowledge with more people online
• Generate steady income from workout plans
• Build a stronger professional reputation

Frustrations:

• Too many disjointed tools for content creation and sales
• Difficulty reaching new clients online
• Time-consuming promotion efforts

How our platform helps: It enables Alex to easily create and share structured workout plans, automatically match with clients, and earn money without needing technical or marketing expertise.

Trainee Persona – Daria, 27 #

Daria is a full-time marketing specialist who often feels overwhelmed by fitness advice online. She’s tried various workout apps and diets, but lacks consistency and often doubts whether she’s following the right plan. Daria wants something simple, trustworthy, and adapted to her busy lifestyle.

Goals:

• Get fit using a plan that fits her personal schedule and needs
• Stay consistent with expert guidance
• Avoid injury and ineffective workouts

Frustrations:

• Conflicting fitness advice online
• Lack of motivation without structured support
• Difficulty choosing the right program

How our platform helps: It matches Daria with the right trainer and plan using AI, provides trusted and personalized guidance, and removes the guesswork so she can focus on progress.

User Stories #

1. As a new user, I want to register as either a trainer or a client so that I can access the features relevant to my role.
2. As a client, I want to fill out a profile questionnaire so that I can receive basic training recommendations tailored to my needs.
3. As a trainer, I want to easily create and upload workout programs so that I can share my expertise without needing technical skills.
4. As a client, I want to browse a catalog of training programs so that I can explore different options.

Initial Scope (MVP) #

✅ The following features will be included in the MVP: #

• User registration for both trainers and clients.
• User profile questionnaire and basic workout recommendations based on input.
• Workout upload system for trainers (support for text, images, and PDFs).
• Content processing algorithm to structure uploaded workouts into standardized cards (with descriptions, structure, and tags).
• Basic recommendation system suggesting workouts based on user profiles.
• Training catalog as a scrollable feed of exercise cards.
• Purchase/save functionality for training programs.

❌ The following functionality is planned for future iterations but is excluded from the initial release: #

• Effortless monetization of programs
• Video pages for trainers
• Real-time messaging/calls
• Mobile app (iOS/Android)
• Social/chat features
• Subscriptions
• Digital products (courses/guides)
• Ratings and reviews

🔗 Repository Link #

https://github.com/IU-Capstone-Project-2025/UrTraining

⚙️ Tech Stack #

Frontend #

• React – Primary framework for building dynamic user interfaces.
• TailwindCSS – Utility-first CSS framework for rapid styling.
• shadcn/ui – Customizable UI component libraries for consistent design.
• Framer Motion – Library for smooth animations and transitions.

Backend #

• FastAPI – High-performance Python framework to build APIs with automatic documentation (Swagger/OpenAPI).
• PostgreSQL – Relational database for structured data storage (user profiles, questionnaires).
• SQLAlchemy – ORM for Python to simplify database interactions.
• Supabase (optional) – Open-source Firebase alternative for authentication and storage.
• Docker – Containerization for consistent deployment across environments.
• CI/CD (GitHub Actions) – Automated testing and deployment pipelines.

AI/ML #

• Sentence-BERT – Pretrained embedding models for semantic search (initial phase, before fine-tuning).
• HuggingFace Transformers – Framework for leveraging LLM (e.g., fine-tuning or zero-shot tasks).
• Faiss – Efficient vector similarity search for recommendations.
• EasyOCR / PaddleOCR – Text extraction from images (e.g., scanned documents or screenshots).

File Storage #

• Cloudinary / Supabase / AWS S3 – Scalable solutions for storing user-uploaded media (images, PDFs).

Weekly Progress Report #

During the first week, our team held several meetings to discuss the core idea of the project and define the roles and responsibilities of each team member. We established the initial scope of the project and agreed on the main technologies that will be used for development.

Challenges & Solutions #

The main challenge we faced was differing perspectives on the project details. To address this, we reached an agreement on the core functionality and defined clear, specific requirements. This ensured that all team members share a unified understanding of the project moving forward.

Conclusions & Next Steps #

During the first week, we made progress by defining the main goal of the project and outlining our future plans. In the upcoming week, our focus will shift to designing the project interface and beginning implementation based on the established backlog.

We confirm that the code in the main branch:

• In working condition.
• Run via docker-compose.