← Back to Projects

Hot Takes

Voice AI for restaurant feedback


Problem

Restaurants rarely get genuine feedback on dishes. Written surveys feel like a chore, and customers either leave generic positive reviews or harsh complaints—missing the nuanced feedback that could actually help improve the menu.

Approach

Designed and built a voice AI agent that engages customers in natural conversations about their dining experience. The system generates QR codes for each table or check, and when scanned, initiates a friendly voice conversation that feels more like chatting with a curious friend than filling out a survey. The conversational format captures nuanced feedback about specific dishes, service, and atmosphere.

Impact

Prototype demonstrated 3x higher engagement rate compared to traditional survey methods. Restaurants using the system gained actionable insights on specific dishes rather than generic ratings.

Technologies

Voice AINext.jsTypeScriptTailwind CSSQR GenerationAnalytics Dashboard