GateFlow
Link to open source: https://github.com/Prashant1567/GateFlow
Link to Live Project: https://ai-builder-lac.vercel.app/
Title: CricLens Command: GateFlow
Overview: Managing foot traffic for a 50,000+ seat stadium during a live match is a massive logistical challenge. CricLens Command is a real-time, agent-driven stadium advisory system built to prevent bottlenecks, ensure fan safety, and optimize gate entry using the power of Gemini.
How it Works (The Agentic Workflow): We moved away from traditional "click-and-update" apps and built a truly autonomous system.
-
Live Ingestion: The dashboard simulates a live feed of unstructured fan messages (e.g., "North gate hasn't moved in 10 mins!").
-
Agentic Processing: The Gemini API continuously monitors this chatter. We engineered a prompt that forces Gemini to act as a "Logistics Engine." It reads the natural language, identifies the specific gate, assesses the crowd density (Low/Medium/High), and calculates the impact.
-
Autonomous Action: Based on Gemini's structured JSON output, the system dynamically updates the live gate capacity meters. If a gate crosses the 85% threshold (Red Zone), the Agent autonomously generates and pins an official rerouting advisory for the fans.
Impact & Innovation: By using Gemini's native language understanding to process messy, real-world social data, we turn panicked fan tweets into structured, actionable logistics data. This prototype proves that AI agents can step out of the chatbox and actively manage physical-world urban flow and safety in real-time.
Tech Stack Built With: React, Tailwind CSS, and the Google Gemini API.
