IPL Knowledge Graph

A real-time IPL Knowledge Graph application that surfaces tactical insights, utilizing Spanner Graph and Vertex AI.

Project Structure

• frontend/: React-based "Mission Control" dashboard that visualizes live match states and tactical shifts.
• backend/: Python-based polling service providing live match updates.
• cf_reasoning_path/: Cloud Functions dedicated to end-of-over tactical analysis.
• hot_path_pipeline.py & setup_graph.js: Core pipeline scripts for ball-by-ball updates and graph management.
• DESIGN.md: Contains information about tokens and design system choices.

Getting Started

1. Set up your GCP credentials by placing an active gcp-creds.json inside the root directory. This explicitly should not be tracked by version control.
2. Initialize backend dependencies and run the backend service.
3. Start the frontend via cd frontend && npm install && npm run dev.