https://drive.google.com/file/d/1QPj9sGfpCQqWtEf9Jb7zkqMZNCCmcqiW/view?usp=drive_link

🎙️ AI LoL Commentator: Real-Time AI-Powered Esports Commentary for League of Legends

Welcome to AI LoL Commentator, an innovative, AI-driven system that transforms ordinary League of Legends (LoL) gameplay into a dynamic, professional esports broadcast. This tool uses real-time data from the LoL Client and Live Client APIs to detect critical in-game events—such as kills, turret takedowns, and objectives—and generates energetic, exciting commentary on the fly.

Powered by Google's Gemini AI for natural language generation and ElevenLabs for high-fidelity text-to-speech (TTS), our system provides an immersive, audio-enhanced viewing experience.

Why This Matters

• For Viewers & Streamers: Adds a professional, engaging audio layer to streams and replays, elevating the viewer experience.
• For Esports: Serves as a prototype for automated commentary, scalable for lower-tier tournaments or dynamic highlight reel creation.
• For AI Enthusiasts: Combines API interaction, real-time event processing, and advanced LLM/TTS technologies into a functional, end-to-end application.

---

How It Works: The AI-Driven Commentary Pipeline

The system operates in a continuous, low-latency loop, monitoring the game state and narrating key moments.

1. Data Acquisition: The program reads the LoL lockfile to securely connect to the local APIs. It then polls the League of Legends Client API (LCU) and Live Client API to fetch real-time game data, including player information and in-game events.
2. Event Detection: The script actively listens for new events as they happen, such as:
   • Game Start: "Welcome, everyone, to the ultimate battleground!"
   • Champion Kills: "Oh, what a play! {KillerName} just took down {VictimName}!"
   • Objective Takes: "The Infernal Dragon has fallen to Team Blue, and the power of the flame is now with them!"
3. AI Commentary Generation: Game events and contextual player data are fed into the Gemini AI. The model, acting as a professional esports commentator, generates concise and impactful commentary.
4. Audio Output: The generated text is instantly sent to the ElevenLabs API for conversion into a high-quality audio file. This audio is then played directly through your system's speakers, providing a truly live commentary experience.

---

Key Features

• Real-Time Event Tracking: Captures and narrates key moments like kills, objectives, and game starts as they happen.
• Direct API Integration: Utilizes the official LoL APIs for a fast, low-latency data stream without external scraping.
• Contextual & Dynamic AI: The Gemini model's chat history ensures that commentary is coherent and follows the game's evolving story.
• High-Fidelity Audio: ElevenLabs provides expressive, low-latency voice output for a professional feel.
• Spectator-Friendly: Operates non-intrusively in spectator mode, requiring no changes to the live game environment.
• User-Friendly Setup: A dedicated ui.py script guides you through the setup process and automatically creates the necessary .env file.

---

Technologies Used

• Programming Language: Python 3.x
• APIs:
  • League of Legends Client API (LCU) & Live Client API
  • Google Gemini API
  • ElevenLabs API
• Libraries: All required libraries are listed in requirements.txt.

---

Setup & Installation

Prerequisites

• The League of Legends client installed and running a game in spectator mode. This is a mandatory step, as the app requires access to the game's lockfile.

Steps

1. Clone the repository:

   git clone https://github.com/rihemar/Seneca_Hacks_Hackathon_2025
   cd Seneca_Hackathon

2. Install dependencies:

   pip install -r requirements.txt

3. Run the setup script: This will prompt you for your API keys and automatically create your .env file for you.

   python ui.py

---

Limitations & Future Improvements

• Latency: Commentary is dependent on the API polling interval. A future improvement would be to integrate with a WebSocket-based system for true real-time event pushing.
• Deeper Context: The current system focuses on major events. Future versions could integrate more nuanced data, such as item builds, gold leads, and player positions, for richer, more strategic commentary.
• Scalability: The current implementation is designed for a single spectator instance. Scaling to support multiple concurrent games would require a more robust backend architecture.

---

License

This project is licensed under the MIT License.