league of leagends smart commentator

Turn any LeagueOfLeagends game into a pro-level broadcast: our AI commentator narrates every detail in real time with natural voice and smart insights.

Comment

Inspiration

As a team of passionate gamers, AI enthusiasts, and software developers, we noticed that many League of Legends matches—especially amateur or community games—lacked engaging commentary. Professional casters bring excitement and insight, but they aren’t always available. We wanted to democratize pro-level commentary, making every match immersive, accessible, and fun for viewers worldwide. This project was inspired by the dataset provided by the Seneca team, which gave us the foundation to analyze real match events and train our AI to understand game dynamics.

What it does

Provides real-time, human-like commentary for League of Legends matches. Narrates every play with contextual insights, from kills and objectives to gold leads and team fights. Uses Natural Language Processing (NLP) + Text-to-Speech (TTS) for engaging, natural audio output. Supports english commentary and can scale to multiple simultaneous games.

How I built it

Game Data Ingestion: Real-time stats for positions, health, gold, kills, and objectives, structured as JSON objects. AI Analysis (NLP): Detects critical events, computes game context, and generates human-like commentary sentences. Text-to-Speech (TTS): Converts commentary into dynamic, natural-sounding audio with adjustable tone and pacing.

Challenges I ran into

Latency: Keeping commentary synchronized with live gameplay. Context Understanding: Capturing strategic subtleties like rotations, zoning, or split-push timing. Naturalness: Making AI speech sound human-like, engaging, and dynamic. Collaboration & Scaling: Coordinating team efforts to handle multiple matches efficiently.

Accomplishments that I'm proud of

Delivered a real-time AI commentator that can narrate every detail like a professional caster. Successfully integrated NLP + TTS to produce natural, human-like voice commentary. Built a scalable system capable of covering multiple matches simultaneously. Leveraged the Seneca dataset to train AI with realistic match events and strategies.

What I learned

How to analyze fast-paced, multi-agent game data in real-time. How to generate context-aware, engaging commentary using NLP. The importance of team collaboration in tackling technical and design challenges. Techniques to optimize TTS systems for dynamic, natural output.

What's next for league of leagends smart commentator

Expand to support other esports titles like Valorant, Dota 2, and CS2. Introduce personalized AI caster voices and avatars. Integrate directly with streaming platforms for on-demand, live commentary. Enhance AI strategic understanding to provide deeper tactical insights in real-time.

Built With

Share this project:

Updates