Read your Taste DNA
Qloo maps who your venue attracts across music, film, fashion and food, then finds the rivals agents compare you with.
AI assistants are starting to choose where people eat, drink and go out. Spotlit shows you how often they choose you, why they pick a rival instead, and fixes it with real taste data from Qloo.
No login for the demo · Built on Qloo Taste AI, A2A and MCP
“Somewhere chill for a date Friday, we're into Khruangbin and natural wine, nothing too pricey.”
Picked Halftone House
Chains have APIs and ad budgets. Independent venues have a vibe that agents cannot read. Spotlit turns that vibe into taste data agents understand, gives your venue its own agent, and tests it against your rivals.
Qloo maps who your venue attracts across music, film, fashion and food, then finds the rivals agents compare you with.
Sixty persona agents built from real taste segments ask a concierge for a night out. It talks to every venue agent over A2A and picks one.
The Growth Strategist explains every lost pick with a reason code and Qloo evidence, applies the fix, and re-runs the arena to measure the lift.
Every line is a real JSON-RPC message between agents. Every score traces back to a Qloo call you can open.
A public A2A agent card and JSON-RPC endpoint per venue, plus an MCP server so Claude, Gemini or any assistant can find venues by taste.
// claude_desktop_config.json { "mcpServers": { "spotlit": { "url": "https://getspotlit.vercel.app/api/mcp" } } } // tools find_venue_for_taste ask_venue get_taste_dna