FanCast

Overview
Language Selection
Teams & Players Selection
View Following Team
Search Result
Latest News

Inspiration

Today, Baseball fans face overwhelming content across 30+ teams and numerous players. Our challenge was to leverage AI technology to help users efficiently discover relevant content while providing a social media-like experience.

What it does

We've enhanced the MLB website with personalized discover features through a new onboarding process:

Greeting & Language Selection

Users choose their preferred language (English, Japanese, or Spanish)

My Team & Player Selection

Users can select their favorite teams and players
Interactive team cards and player selection
Search functionality for quick access
Optional step - users can skip

Discover Page Features

AI-powered content digest for favorite teams including: - Key Game Results - Monthly Analysis - Division Race Implications - Current Division Standings - Schedule
Multimodal content of highlight clips and articles
Search functionality for exploring new teams/players
"Follow" button to modify team preferences

Technical Implementation

Frontend: React.js
Backend: Node.js, Gemini AI
Database: Firebase
Features:
- Real-time, multimodal content integration
- Personalized game summaries
- Dynamic content feed
- Performance-optimized architecture

Challenges

Balancing UX ambitions with technical feasibility
Multi-language content translation
Optimizing load times for digest content

Key Achievements

Intuitive onboarding flow with team/player selection
Efficient keyword-based search functionality that works for both teams and players
Seamless integration with MLB's existing digital experience
Real-time content curation using Gemini AI

Learnings

Gemini AI Integration:
- Effective processing of multi-source content
- Dynamic summary generation
- Context-aware content delivery
Performance Optimization:
- Balanced real-time updates with system performance
- Implemented caching strategies
- Optimized batch requests
- Used lightweight data formats