Understanding Amplifier AI’s Game Recommendations API: How to Integrate and Optimise AI-Driven Player Engagement
Overview
Amplifier AI’s Game Recommendations Engine delivers personalised game recommendations that drive player engagement and broaden game preferences, directly contributing to increased revenue for casino operators.
Recommendation Categories
The Game Recommendations Engine provides various types of recommendations to meet different use cases, with each category tailored to serve a specific carousel or feature on-site.
The categories and structures available from the API are shown in the swagger documentation, and explained below:
Future-Anthem Personalise (requires API key)
- Recommended For You: Top recommendations for each player, from across Amplifier AI’s various models
- New Games: New games re-ordered for each individual player, based on similarity to their favourite games
- Because You Played:
- Recommendations from this API are based on the similarity to a player’s favourite games, from Amplifier’s game similarity model, which utilises game meta-data
- For each set of recommendations, the favourite game on which they are based is provided as "context" in the response
- In the below example, game id 000049 is the player’s favourite which the 'Because You Played' recommendations are based on, hence can be used “because you played game 000049……”
Where a player has multiple favourite games, this category could be used to drive multiple carousels or sections - Players Like You: Recommendations based on players who have similar favourite games from Amplifier AI’s collaborative filtering model
- Game Discovery: Games that are different to what the player normally plays. This is included to allow a ‘try something different’ type carousel
- Popular Games: Recommendations of the most popular games on-site
- Trending Games: Recommendations based on games that are showing the biggest increase in activity
- Default APIs:
- New Players Recommendations: Recommendations for new players which are optimised against best converting games. These will be returned automatically if recommended_for_you is called with an unknown player id, or a player id with no significant activity
- New Players Conversion: Personalised recommendations for players in early lifecycle – these will automatically be returned if recommended_for_you is called for a player in early lifecycle
-
- Other categories in the default section map to the ‘recommendation’ section, and will be returned where the player id is unknown
Attributes (Your Favourites):
- A player’s favourite games based on Amplifier AI’s significant activity model
- Typically, these would be used instead of ‘recently played’, as a more accurate version of a player’s recent favourites
Site Personalisation:
- It is recommended to call categories independently and employ lazy loading
- However, this category will consolidate all categories into a single payload
If a player has no transaction history, they receive the new player recommendations. These are non-personalised recommendations (i.e. all new players get the same recommendations) that are designed to active brand new players who have no prior activity.
Recommendations can also be customised based on data availability to ensure compliance with local jurisdictions and game type specifications (e.g., slots-only recommendations where applicable).
For more information:
- Discover how Game Recommendations work.
- Learn how we measure performance to showcase ROI and other key metrics.
- Refer to our Integration Requirements Guide for Amplifier AI Game Recommendations integration.
Still have questions? Check out our FAQs for further assistance.