-
Live Qloo results: a fair restaurant shortlist for two very different tastes.
What it does
Common Table turns the daily “where should we eat?” conversation into a fair, taste-aware group lunch decision. Two to four people add a city and a few things they love. Common Table finds actual restaurants that Qloo recommends for everyone at the table, then ranks the fairest shared picks first.
Why it matters
Lunch is a decision teams repeat almost every workday. Generic nearby lists ignore taste, and group chats tend to let the first suggestion win. A good group pick should work for the people going—not just the person who spoke first.
How Qloo powers it
Common Table resolves each diner's cultural taste anchors with Qloo Search, then calls Qloo Insights separately for each person with restaurant, city, and explainability filters. A restaurant only qualifies if the same Qloo place entity appears in every person's results. The order balances individual Qloo affinities, so one strong score cannot hide a weak match for someone else. Results show each person’s score and useful Qloo venue tags.
Agentic by design
The live app exposes find_shared_lunch as an MCP Streamable HTTP tool at /mcp. An assistant can pass a city and two to four taste profiles and receive structured restaurant picks, per-person affinities, and map links—the same live planner used by the web app.
Built for a workday
Remember a regular lunch table on one device, then use it again when the next “where should we go?” message appears. This setting is off by default and can be enabled by the user. Maps links search the restaurant's exact name and address; verify current opening hours and other details with the restaurant.
Try it
Open the live demo and choose “See a live example” to run the prepared Bengaluru pair. Or enter your own city and taste anchors. Qloo results vary with the selected area and signals.
Technology
Qloo Taste AI API, MCP Streamable HTTP, Node.js, HTML, CSS, JavaScript, and Vercel. The Qloo key is stored server-side and is not included in the public source. The repository includes setup instructions and an MIT license.
Built With
- css
- html
- javascript
- model-context-protocol
- node.js
- qloo-taste-ai-api
- vercel