The difference

Real-time personalization and recommendation

We marry enterprise internal data with external data to deliver personalized recommendations in real-time.

Add new categories and merchant on-the-go

A living, breathing map of the world’s tastes

A living, breathing map of the world’s tastes acquires essential data on various lifestyle categories, to create a graph-based entity-affinity model.

This model is then mapped to customer behavior data, to reveal a universe of choices and recommendations.


The TasteGraph™ has over 6.5Mn merchants analyzed and mapped, including








other merchants


With over 6.5Mn merchants and counting, every digitally discoverable merchant is a target entity for the TasteGraph™. It identifies unique merchants globally and establishes relationships with other relevant merchants.

understands how customers make decisions

Taste Attributes understands how customers make decisions in each lifestyle category. It identifies the attributes that define tastes in these categories. For example, to understand the tastes of customers in dining, factors restaurant attributes like quality of service, ambiance, food, cuisine and specials. With cutting-edge machine learning and natural language processing (NLP), filters through a combination of structured metadata and unstructured data. It then tags

every merchant with hundreds of such taste attributes and assigns scores to each, to create a taste profile of the merchant.


Based on a merchant’s profile, provides an affinity score between any two merchants in each category. uses anonymized customer preferences and a patented collaborative method to create a cross-category graph.
All in real-time.

Complement your customers’ tastes with the right merchants

TasteMatch score is used to rank recommendations for customers

The TasteMatch provides an affinity score between a customer’s taste profile and a given merchant.

The TasteMatch score is used to rank recommendations for customers. It considers context for specific use-cases.

The math behind simple choice

personalization apl

Consumer choice relies on four components: taste, influence, context and behavior. uses these components to understand choice. We call it the Choice Equation.

Aggregated preferences of customers, like ratings and reviews

External social interactions that affect a customer’s decisions, such as ‘likes’ and ‘shares’

Other factors such as devices, time,
location, weather and more

Online and offline customer activity such as past transactions and interactions

Patented algorithms

Combines multiple recommender systems to deliver the top affinities in any category to every customer.

Cognitive models

Is underpinned by an extensive set of dimensions that mimic how consumers make choices.

Iterative machine learning

Evolves constantly in a built-in machine learning loop.



Aggregate Rating
3 based on 29 votes

Speak to our AI-personalization experts today

Stay on top of your game in the age of relevance


Check out our blogs, articles and more


Discover what's new in the big data and AI industries


Keep up with the latest updates from