← Project index
RecordAI · Spatial interaction
Glo AI Chatbot & Motion-Interactive Wall
Led the delivery of an AI-driven, motion-interactive installation combining conversational AI with Kinect-powered visuals for live brand activations.
- Organisation / Context
- Globacom
- Role
- Lead engineer
- Timeline
- 2 months
- Date
- February 2025
Overview
Architected and delivered an immersive installation for Glo that fused real-time conversational AI with Kinect-based motion tracking. Users interacted verbally with an on-screen assistant while their physical gestures dynamically controlled large-scale visual effects. The system was designed for high-footfall environments, prioritising responsiveness, stability, and intuitive interaction. It successfully blended AI, physical movement, and storytelling to create a memorable brand experience.
- Outcome
- Responsive AI and motion interaction for a live brand activation
Engineering
- 01Used the Kinect SDK within Unity to read body-tracking input and drive the motion-reactive visual experience.
- 02Delivered the AI chatbot as a separate web experience on another public display.
- 03Turned the visitor's phone into a remote keyboard, using bidirectional communication to keep the phone, AI system and display synchronised.
- Hardest problem
- The installation combined different interaction surfaces without making the experience feel fragmented. Body movement drove the Unity visuals, while phone input and AI responses moved between a separate web experience and public display in real time.
- Stack
- Unity · Kinect SDK · C# · Node.js · WebSocket · AI
Media
Project views
02
