TV-sized awesomeness

About the Project

🚀 Inspiration

The inspiration behind this project was to reimagine the living room experience using AI and cloud power — combining the simplicity of remote navigation with the intelligence of dynamic content. We wanted to build an app that feels like Netflix met ChatGPT, on a smart TV — but without needing a single app store download.

🧠 What We Built

We built a 1920x1080 full HD web app designed specifically for the Synamedia Senza platform. It features:

• Multi-screen navigation using just arrow keys, Enter, and Escape
• AI-powered menus (simulated using async dynamic loading)
• Nested navigation, letting users dive deep into categories like “AI Picks”, “Top Trends”, or “Recommended Just for You”
• A slick and intuitive UI optimized for TVs and remote control
• Lightweight, single-file HTML with no external frameworks — Senza-ready by default!

🛠️ How We Built It

• Pure HTML5, CSS3, and vanilla JavaScript
• Dynamic content rendered with simulated API calls (can be swapped for OpenAI/LLM backends)
• State-managed screen switching (menu → details → back)
• Designed and tested at full 1920x1080 resolution for seamless compatibility with the Senza cloud connector
• Focused on snappy navigation, clarity, and scalability

// Sample screen switch logic
function goToScreen(screenName) {
  hideAllScreens();
  document.getElementById(screenName).style.display = 'flex';
}

💡 What We Learned

• How to design for TV-sized UIs with low-latency remote inputs
• How to build navigation systems that work without mouse/touch
• How to think like a product team building for 10-foot experiences
• How to simulate AI-powered UX in a constrained environment — and still make it feel magical

😅 Challenges We Faced

• Designing a UI that looked good at 1920x1080 without overwhelming the viewer
• Handling keyboard-only input logic elegantly and smoothly
• Keeping everything performant in a cloud-streamed browser environment
• Ensuring that navigation between screens was clear and intuitive
• Fitting all this into one hour of build time — challenge accepted and conquered 💥

Built With

• css3
• html5
• obsstudio
• openai
• vanilla
• vsc