When attending a World Cup match or a massive stadium concert, the excitement often fades when reality hits: 30-minute bathroom lines, chaotic entry gates, and terrible crowd congestion.
For the PromptWars Hackathon hosted by Hack2Skill and Google Cloud, we decided to tackle this multi-million dollar logistical nightmare head-on.
Enter** VenueIQ **a dual-panel, serverless smart venue platform entirely orchestrated by Real-Time Data and Autonomous AI Agents. 🤯
🔗 Live Demos & Code
🚀 Live Cloud Run Platform: https://venueiq-frontend-85430108611.us-central1.run.app/
💻 GitHub Open Source Repo: https://github.com/dev200413y/VenueIQ.git
(Our live demo is currently connected to a synthetic data-stream — you can literally watch crowds move and wait-times shift in real-time right in your browser!)
🧠 The Core Engine: Multi-Agent AI Architecture
Instead of just building a dashboard full of static charts, we embedded two distinct, autonomous AI personas into the platform's core memory using Mistral AI via the Groq Inference Engine.
**
1. The Operations Agent (For Ops / Admins)** 🕵️♂️
This agent lives inside a futuristic, 6-tab Admin Command Center. It continuously ingests realtime density metrics from Firebase. It doesn't just display a graph if there's a stampede risk at the "North Gate" — it executes predictive logic, flashes critical alerts, and automatically generates crowd-rerouting strategies for on-ground deployment teams!
- *The Personalized Virtual Concierge (For Attendees) *📱 We built a mobile-first app that tracks the exact seat coordinates of the fan. When the fan types, "Where is the nearest restroom with the shortest wait?", the AI cross-references the live facility database to give them dynamic, logic-based directions. It guarantees they get back to their seat before the next half starts.
🔥 Mind-Blowing Features We Built
We went all-out to guarantee this platform felt like a $1M enterprise product:
✅ AI Camera Feeds What if CCTV cameras go offline? We solved this by generating a high-performance SVG + Canvas rendering engine. VenueIQ dynamically draws hundreds of animated, expanding nodes representing live humans based purely on raw integers bouncing across the database payload!
✅ Live Facilities Wait-Time Monitoring Attendees no longer have to guess. VenueIQ pulls massive amounts of data to display literal wait times and progress bars for individual Restrooms, Food Courts, and Merch Stalls, instantly updating on their mobile app via WebSockets.
✅ Real-Time Revenue & Staff Synchronization Ops managers get granular control! With highly stylized sliders, admins can literally drag and re-allocate security personnel across 12 distinct stadium zones. Every move pushes state to Firebase in <100ms.
✅ Dynamic Queue Intelligence A dedicated panel where the Mistral AI monitors average gate wait times, automatically flagging bottlenecks (e.g., "Gate C is at 94% capacity") and autonomously offering load-balancing alternatives.
🛠️ The Tech Stack (Serverless & Lightning Fast)
Frontend Magic: React 19 + TypeScript powered by Vite. (Because speed is everything).
Database state: Firebase Realtime Database. (No slow polling — pure WebSocket magic for <100ms sync times).
AI Engine: Mistral AI executing high-speed inference queries.
Hosting Deployment: *We bypassed standard hosters and fully containerized the platform via a multi-stage Dockerfile utilizing Nginx, deploying directly to Google Cloud Run for dynamic, auto-scaling performance.
*🚀 Conclusion
The biggest lesson from this hackathon was learning how to stop using AI simply to "generate text." By pushing AI to strictly orchestrate logic and routing, we created an autonomous ecosystem. Implementing a serverless architecture allowed us to execute the entire operations dashboard maintaining massive cost-efficiency on Google Cloud Run.
Massive shoutout to @googlecloud, and the PromptWars organizers for hosting this incredible event! 🙌
If you like what we built, drop a comment, give the GitHub repo a ⭐, and follow along. This is just the beginning of making mega-event experiences enjoyable again.
Top comments (0)