Customer orders, product data and customer application clickstream data flow through AmazonRedshift, Amazon Relational Database Service (Amazon RDS), Amazon EMR, Amazon Simple Storage Service (Amazon S3), and Amazon Kinesis. Data is ingested, transformed, and enriched using AWS Glue, Apache Spark on Amazon EMR, AWS Lambda, or any number of other means. Bulk and batch data are loaded to Neo4j using the Neo4j Spark Connector for Amazon EMRor Neo4j APIs (including Cypher and Arrow) running in Lambda. Near real-time data from Apache Kafka on AWS by Confluent flows through the Neo4j Kafka Connector to the Neo4j database. Transaction events from Neo4j can also be pushed to Kafka. Clickstream data from Amazon Kinesisflows through the Neo4j Spark Connector running on Amazon EMRto the Neo4j database. The Neo4j Graph Database (GDB) and Neo4j Graph Data Science (GDS) let you store, query, analyze, and manage highly-connected data. Neo4j Aura is deployed as a fullymanaged service on AWS. Neo4j Enterprise runs on Amazon Elastic Compute Cloud (Amazon EC2) infrastructure instances. Data scientists create features on Neo4j GDB and GDS. Embeddings are exported to Amazon SageMakerfor smarter ML. Data scientists can also use the Spark and Python APIs to implement complex ML on Neo4j. Data scientists can push graph features back to relational source systems where they can be easily visualized or reanalyzed. Amazon QuickSightand Neo4j Bloom visualization tools let analysts explore data and present findings plainly. Developers use official and unofficial drivers to access Neo4j using Java, Node, Javascript, C#, Python, Go, Ruby, PHP, Erlang, and Perl from AWS Lambdaor any application. Others choose a low-code approach using GraphQL API and the GRANDStackframework, using Amazon API Gateway, AWS CodeDeploy, and other development services. The recommended services using Neo4j GDB and GDS run fast, at scale, in near real-time. This framework can be used to build similar recommendation engines fo
