Recommendations
Personalized recommendations start with Qloo’s predictive Taste AI™
Qloo® combines an extensive catalog of global cultural entities with a deep understanding of consumers’ interests across those notable people, places, things, and lifestyle interests.
Personalized recommendations start with Qloo’s predictive Taste AI
How Taste AI powers personalized recommendations
Client CDP
Client Input
- /transaction_history
- /customer_lat_long
- /customer_gender
- + others
Qloo AI API
Qloo Output
- /brand_affinities
- /entertainment_affinities
- /destination_affinities
- + others
Client Interpretation
Customer UI
Basic examples
Here's how to get restaurant recommendations based on known dining preferences:
Bash
curl --request POST \
--url 'https://staging.api.qloo.com/v2/insights/' \
--header 'X-Api-Key: YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data '{
"filter.type": "urn:entity:place",
"signal.interests.entities.query": [
{"name": "Nobu", "address": "105 Hudson St, New York, NY"},
{"name": "Balthazar", "address": "80 Spring St, New York, NY"}
],
"filter.location.query": "Chicago",
"take": 5
}'filter.type- Set to urn:entity:place to scope results to restaurants/places
signal.interests.entities.query- Array of known restaurants (with name & address) to infer taste
filter.location.query- Where you want recommendations returned (independent of signal locations)
take- Number of results to return
Signals in one domain can drive results in another — here, music and fashion taste return places:
Bash
curl --request POST \
--url 'https://staging.api.qloo.com/v2/insights/' \
--header 'X-Api-Key: YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data '{
"filter.type": "urn:entity:place",
"signal.interests.entities": [
"urn:entity:artist:brian-eno",
"urn:entity:brand:jil-sander"
],
"filter.location.query": "Kyoto",
"take": 5
}'signal.interests.entities- Entity URNs from any domain — artists, brands, films, destinations
filter.type- The domain you want back, which need not match the signals
filter.location.query- Optional geography to scope the results to
take- Number of results to return
Pass every member's signals in one request to find what the whole group responds to:
Bash
curl --request POST \
--url 'https://staging.api.qloo.com/v2/insights/' \
--header 'X-Api-Key: YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data '{
"filter.type": "urn:entity:place",
"signal.interests.entities": [
"urn:entity:artist:brian-eno",
"urn:entity:movie:lost-in-translation",
"urn:entity:videogame:journey"
],
"filter.location.query": "New York",
"take": 10
}'signal.interests.entities- Combined signals across the group; affinities are weighted together
take- Raise this for a longer shortlist the group can choose from
Practical implementation
Curate content with precision
From the latest blockbusters to niche articles, interpret your customers' diverse tastes and recommend the best content in return.
Suggest social events for individuals and groups
Help customers navigate an ocean of options by displaying events and activities that are hyper-relevant to their interests.
Personalize individual travel itineraries
Travel recommendations that align perfectly with personal preferences, increasing conversion and satisfaction.
Create targeted offers that connect
Acquire new customers and keep existing ones loyal by promoting offers, rewards, and deals that hit the mark every time.