For judges

Every prepaid store runs the same table outside. Stoop plans one for the community around it, with Qloo's taste data deciding the music (from the city) and the food spots, prizes and venues (from the places near the store).

The path, in 6 steps

Open the example event
  1. 1

    Same table. Different community.

    On the landing, three example stores with the same kind of deals get three different plans: different music, food spots and prizes, each picked by Qloo for that city and the places around that store.

    See the three stores
  2. 2

    One event and its receipts

    18th Street Wireless, Chicago. Tap "Why Qloo picked this" on the playlist, a food spot or a gift card: the exact Qloo request behind it (never the API key).

    Open a receipt
  3. 3

    The generic AI plan beside it

    The same model, the same inputs, no Qloo. It names no place near the store, and none of its 3 artists make Qloo's own top 20 for Chicago adults 25 to 34.

    Open the comparison
  4. 4

    The first real table

    Oct 30, at the founder's store in Kennesaw, GA: the first table Stoop planned, with a photo, the drop board and what the store saw that day, labeled as one day.

    See the real table
  5. 5

    Take it to where people go

    On the Houston example, Qloo's parks and venues near the store have "Plan it here", and the event has "Get the OK": drafts to the district manager and the organizer.

    Open the drafts
  6. 6

    Try your own store

    Type a store's name and address, see its community, then plan the event in about a minute. If live plans are paused, Stoop says so and opens the nearest example store.

    Plan an event

What Qloo decides

Each Qloo call and what it decides
CallDecidesCode
/v2/insights · artistsignal.location.query = the city, signal.demographics.age = the store's adult age bandsThe playlist: the artists the city ranks highest for those ages.lib/plan/requests.ts
/v2/insights · placefilter.location = the store's POINT, filter.location.radius = 1500 (about 1 mile), one category per call, explainabilityThe community: food spots, cafes, venues and parks near the store, each with Qloo's reason. The food spot, the gift cards, any partner and any venue for the table come only from these.lib/plan/taste.ts
/v2/insights · artist by placessignal.interests.entities = venues in the communityA second music option: what people who love the community's top local spots play. Offered only when it is current and adds artists the city's list lacks, and always next to the city's music.lib/plan/block-music.ts
/search, then /v2/insights · scoringthe generic plan's artist names, then one filter.results.entities call against the city's own picksRanks the generic plan's artists against the city's own picks, all in one call, so the comparison is fair.lib/plan/baseline.ts

The city sets the music; the places Qloo finds near the store set the food, the gift cards and the partners. In our tests a neighborhood point barely changed Qloo's taste results, so Stoop never claims a neighborhood's taste, and it never treats a shop's own Qloo profile as the taste of the people who go there.

Honest by design

  • Every music, food and venue pick has a receipt, or says it isn't from Qloo.
  • Every place and artist named in a plan comes from that store's own Qloo results.
  • Deals are quoted word for word, as the manager typed them.
  • Stoop writes the posts; the manager shares them. Nothing goes out on its own.
  • Example stores are fictional. No carrier logos.
  • One real result, from one day, labeled as a first look, not proof. The pilot keeps counting at both stores.

Also look at