Letting Gemini Write SQL Against BigQuery, So Fans Don't Have To
Personalized baseball coverage that actually personalizes: follow your players, pick your schedule, and an LLM-translated SQL layer lets the AI generate stats and visualizations against fresh MLB data without code changes.
Sports coverage is bimodal: encyclopedic if you're a stathead, drive-by if you're casual. There's almost nothing for a fan who wants depth on their three favorite players and a glance at everything else. The Google MLB Hackathon was a chance to build that.
Natural language → SQL → chart
Gemini reads the user's question and the BigQuery schema, emits a SQL query, runs it, and hands the result to a chart generator. New MLB datasets land in BigQuery and the system can answer questions against them without a single code change. The whole loop is multilingual — English, Japanese, Spanish — including the article output and PDF exports.