mma-coach-assistant

An AI-powered MMA coach that analyzes fight videos and generates personalized training plans using Google Gemini.

Comment

Inspiration### Explanation of Video 2: "MMA Coach Assistant" Demo

This video demonstrates the user experience of the MMA Coach Assistant, an AI-powered web application designed to analyze MMA fight footage and generate actionable insights, training plans, and personalized merchandise recommendations. The video showcases a complete workflow from input to output, highlighting the integration of multimodal AI (video + text) with e-commerce.

1. Initial Setup & Input (0:00–1:13)

  • Interface: The app opens with a clean, dark-themed UI titled "MMA Coach Assistant".
  • Fields:
    • Fighter Name: User enters Nassourdine Imavov.
    • Opponent Name: User enters Caio Borralho.
    • Weight Class: Changed from "Lightweight" to "Middleweight".
  • Upload Section: A drag-and-drop zone prompts users to upload a fight video in formats like MP4 or MOV (max 100MB).
  • Action: The user selects a file named "Imavov vs Borralho | HIGHLIGHTS | UFC Fight Night.mp4" (74.15 MB), which appears in the upload area with a green checkmark.

✅ This step shows user-friendly form handling and file validation.

2. Analysis Process (1:14–2:28)

  • After clicking "Analyze Fight", the app initiates a two-phase process:
    • Upload Progress Bar: Shows the video uploading (progresses to 50%).
    • AI Processing Indicator: A spinning loader appears with the message:
      > "AI is analyzing the fight..."
  • The interface remains static during processing, indicating backend computation using Google Gemini AI to extract insights from the video.

🔍 This phase leverages multimodal AI to analyze visual content (fight footage) and contextual text (fighter names, weight class).

3. Results Display (2:29–2:51)

Once processing completes, the results appear:

Performance Metrics
  • Strike Accuracy: 45%
  • Successful Takedowns: 0
  • Avg. Strikes/Min: 3.4
Key Insights
  • Imavov used precision striking and head shots effectively.
  • He had impeccable takedown defense (blocked all 5 attempts).
  • Showed excellent octagon control and strategic patience.
  • Opportunity to integrate more offensive grappling.
Recommended Training Focus
  • Develop advanced combination drills (head-body-leg).
  • Add offensive wrestling and Judo throws.
  • Improve defensive footwork and head movement.
Weekly Training Schedule

A full 7-day plan is generated:

  • Monday: Technical Striking & Flow Drills
  • Tuesday: Offensive Grappling & Takedowns
  • Wednesday: Strength & Conditioning
  • Thursday: Clinch Work & Body Striking
  • Friday: Live MMA Sparring
  • Saturday: Active Recovery
  • Sunday: Rest Day

🎯 This section demonstrates AI-generated, data-driven coaching — transforming raw video into a structured training program.

4. Integrated E-Commerce (2:52–3:11)

The app transitions to a merchandise store for Nassourdine Imavov:

  • Store Page: Displays "The Official Nassourdine Imavov Collection" with a hero image placeholder.
  • Featured Products:
    • Fighter Tee ($45.00)
    • Fighter Hoodie ($75.00)
    • Fighter Hat ($35.00)
    • Signed Poster ($150.00)
  • User Action: Adds multiple items to the cart.
  • Cart Summary:
    • Subtotal: $230.00
    • Options to "Proceed to Checkout" or adjust quantities.

🛒 This highlights AI-powered product recommendations based on the fighter’s profile and performance — a seamless blend of sports analytics and commerce.

✅ Summary: What This Video Shows

Feature Description
Multimodal AI Uses video analysis (Gemini) + text input to understand fight dynamics
Real-Time Feedback Upload → Processing → Instant insights
Personalized Coaching Generates metrics, key insights, and a weekly training schedule
E-Commerce Integration Recommends official gear based on the analyzed fighter
User Experience Intuitive, responsive UI with clear feedback at every stage

💡 Key Takeaway

This video proves that AI can transform combat sports preparation by:

  • Automating time-consuming film study
  • Providing objective, data-backed analysis
  • Delivering personalized training plans
  • Creating monetization opportunities through fan merchandise

It’s not just a demo — it's a complete product vision for the future of MMA coaching and athlete branding.

🔗 Related Links

Let me know if you'd like a shortened version for social media or a presentation slide deck!

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for mma-coach-assistant

Built With

Share this project:

Updates

posted an update —

graph TB A[React Frontend
TypeScript + Vite] --> B[API Gateway] B --> C[Google Gemini AI
2.5 Flash] B --> D[Cart Service
C# + Redis] B --> E[Product Catalog
Go] B --> F[Recommendation
Python + AI] B --> G[Payment Service
Node.js] B --> H[Email Service
Python] B --> I[Shipping Service
Go] B --> J[Currency Service
Node.js] B --> K[Checkout Service
Go] B --> L[Ad Service
Java]

D --> M[(Redis Cache)]
E --> N[(Product DB)]

style A fill:#61dafb
style C fill:#ff6b6b
style D fill:#239120
style E fill:#00add8
style F fill:#3776ab

Log in or sign up for Devpost to join the conversation.