Oracle outlined a handful of MySQL HeatWave advances with the most notable being Lakehouse availability on Amazon Web Services as well as innovations that make the platform more autonomous.

The company announced the MySQL HeatWave new capabilities at Oracle Cloud World in Las Vegas. MySQL HeatWave is a database managed service that enables developers to create and deploy applications on the open-source database. HeatWave, which is available on Oracle Cloud Infrastructure, is an in-memory query accelerator that speeds up analytics on MySQL data without the need for Extract, Transform and Load (ETL) operations.

Some of the use cases for MySQL HeatWave, which competes with Google BigQuery, Snowflake, AWS Aurora and Redshift and similar platforms, include digital marketing campaign monitoring and customer classification, gaming personalization, healthcare analytics and fintech fraud detection and portfolio analysis.

Constellation Research analyst Holger Mueller said the Oracle MySQL Heatwave capabilities could woo developers. He said:

"The HeatWave team has out innovated the cloud database industry by bringing together 5 key innovations, and most recently adding lakehouse support. And of course, HeatWave doesn’t stop here – with availability of Lakehouse on AWS, the key addition of vector format for including transactional data for LLM models, better and more AutoML in the database and becoming a more and more autonomous database thanks to Autopilot advancements."

More from Oracle Cloud World:

Here's a look at the key MySQL HeatWave additions:

  • HeatWave Lakehouse, which scales MySQL HeatWave across multiple object stores and data sources, now supports Apache Avro file format and multiple compression algorithms. HeatWave Lakehouse now supports 3 file formats including CSV, Parquet and Avro.
  • The addition of Avro support enables HeatWave Lakehouse to connect to key platforms such as Kafka, OracleDB, Azure and Google BigQuery.
  • MySQL HeatWave is available in Oracle Cloud, AWS and Azure. Oracle said the addition of HeatWave Lakehouse on AWS now combines 5 AWS services into one.
  • HeatWave AutoML is supported with HeatWave Lakehouse for machine learning training, inference and explanations in data object store for automated training that keeps models up to date. Heatwave AutoML competes with AWS Redshift and Snowflake's SnowPark. Oracle said HeatWave AutoML, which now supports Tfldf and BERT for text processing, is designed to remove manual model steps and reduce the need for data scientist expertise.
  • Generative AI with MySQL Heatwave vector store so users query and retrieve information in natural language to handle documents in various formats. Use cases include retrieval, summarization, code generation and classification.
  • Workload aware machine learning automation including Autopilot indexing new Auto Unload, Auto Compression and Adaptive Query Execution.
  • JSON support in MySQL. MySQL uses binary JSON representation in storage with availability in various languages. Oracle said the new analytics capabilities in MySQL Heatwave line up with capabilities found in Snowflake and Databricks in a head-to-head comparison.
  • JavaScript supports in-database programming as well as tools to bulk ingest data faster across multiple clouds. Oracle added multiple capabilities to ingest data from AWS data stores.