Reference Architecture

Reltio Data Science seamlessly connects the data managed within the Reltio Connected Cloud to Apache Spark and thereby quickly enables agile and closely predictive analytics.

This not only provides for faster time to analytics but more relevant and accurate information that companies can rely on.

In the Reltio Data Science environment, the data flows run in a closed-loop, from the Reltio Connected Cloud through Amazon S3 to Apache Spark. Here, the data is processed for analytics and is sent back to the profiles as analytics data available for the Reltio Connected Cloud users.

What You Can Do With Reltio Data Science

You can export entities and relationships from your Reltio tenant into Reltio Data Science, along with the interactions data from the enterprise. The platform thus provides a foundation data set, with all information in one place. Data Analysts, Data Scientists, and Business Users can now create analytical and predictive models on top of Reltio Data Science. They can even bring in their own analytical algorithms by importing them into Spark and applying them against combined data.

Interactions are designed to represent high-volume data that is critical to run analytics. Examples of interactions are customer call activity, claims transaction, prescription data, script data, sales transaction, and so on.

Once you run analytics on the Reltio Data Science platform, you can save the aggregated data into the Core platform using analytics attributes, such as total script volume, total customers in a region, and so on.

Benefits for Your Business

With Reltio Data Science, you can improve efficiency and reduce your data management efforts by using:

  • SDK for the initial load and ongoing updates (no need to extract data from your MDM and load it into S3 or DW)
  • Integration with Reltio Data Science and Spark Provider with built-in auto scaling (faster way to spin up and manage servers to run the application)
  • Reltio Data Science SDK and Interactive Notebooks (robust toolset to develop and migrate your analytics)
  • SDK to push data back to Reltio MDM (ready-to-use tools to save results and update data back in master profiles)
  • Automatic data syncs (no efforts needed for bi-directional synchronization)

To provide reliable data for analytics and Machine Learning (ML), Reltio puts all data that you need for analytics and ML in one place, to be accessible in real-time. Reliable and accurate data from master profiles, interactions, third-party, public and social media sources is consolidated for deeper analytics. When analytics run on a data foundation, organizations can make better and informed decisions.

Another benefit is a closed-loop between intelligence and action. Reltio Data Science brings aggregated analytics back to master data profiles to enrich and improve the data. Unlike analytics-only tools, Reltio's bi-directional connectors support bringing back aggregated insights to master data profiles to users of data-driven applications.

This is all done at the speed of business. Faster correlation of accurate, up-to-date profile data with transaction data from multiple sources makes deployments quick and simple. Pre-configured connectors to analytics environments make leveraging reliable data easy. Data is made available to the analytics apps and ML algorithms in near real-time.

In a traditional approach, to drive insights, you need to design several steps to extract master data out of your MDM application and load interactions into the analytical environment and create a process to apply the ongoing changes. Reltio Data Science automates the initial ingestion of entities, relations, and interactions from S3 into Apache Spark and provides the ability to save calculated results back to Reltio MDM in the form of analytics attributes.