KraowKao

[ Overview ]

KraowKao is a polyglot feedback platform designed for hospitality businesses in Bangkok. It lets customers leave structured reviews via QR code at their table — positive reviews get gently nudged to public platforms (Google, TripAdvisor), while negative feedback stays private so the business can fix issues without public shaming. The name comes from Thai: เกรียว kao (ripple) and เกรียว kao (fame) — a ripple of reputation. Built with three languages that each do what they do best: Django handles the feedback intake pipeline, Elixir powers real-time notifications, and Next.js delivers the dashboard.

[ Key challenges ]

• • Polyglot architecture coordination — Django webhooks feeding into Elixir GenServers for push delivery
• • QR code generation and table mapping at scale across dozens of restaurant locations
• • Smart review routing — sentiment analysis determines public vs. private path without false positives
• • Real-time push notifications to restaurant staff via WebSockets when negative feedback comes in

[ Key features ]

• • Anonymous structured feedback submissions via QR code at each table — no app download needed
• • AI sentiment analysis routes positive reviews to public platforms, negative feedback stays private
• • Real-time dashboard with push notifications — staff sees issues the moment a customer submits them
• • Analytics on customer satisfaction trends, common complaint categories, and resolution tracking
• • Weekly email reports with actionable insights and improvement recommendations
• • Multi-language support (English/Thai) for Bangkok hospitality staff and customers

[ Screenshots ]

[ Tech stack ]

• • Django backend handling feedback intake, sentiment analysis pipeline, and business configuration
• • Elixir/Phoenix for real-time WebSocket notifications and push delivery via GenServers
• • Next.js frontend for business dashboard, customer feedback interface, and analytics
• • PostgreSQL for feedback records, business profiles, and aggregated analytics
• • Redis for pub/sub message broker between Django ingestion and Elixir push delivery
• • AWS deployment with S3 for QR code assets and Docker for polyglot service orchestration