PoolParty Recommender in Action

This section will attempt to give you initial insights into the individual steps of the recommendation process along with a brief illustration of the creation of a semantic footprint and finally delivery of the results.

To recap, the recommendation process is comprised of the following steps:

The superlative feature of the PoolParty Recommender is that the entire process is clearly comprehensible, traceable and transparent. You can easily follow the reasoning from the input all the way to the result. In our approach the user can see why a particular result was returned by the system. Another outstanding feature is the combination of concept-based search by the PoolParty Suite with the Elasticsearch full-text search functionality making the PoolParty Recommender a future-proof system. This allows subsequent integration of machine learning or large language models maintaining the provision of traceable recommendations being an essential foundation for the corporate decision making process.

The PoolParty Recommender solution is comprised of the API based PoolParty Middleware Recommender and the PoolParty Recommender Workbench to set up and test the recommender configurations. Its API based architecture provides for easy integration with any existing frontend solutions.