- Oracle unveiled new AI features for its HeatWave MySQL database service at Oracle CloudWorld.
- Enhancements reduce the need for manual integrations, AI expertise, and complex security setups.
- HeatWave on AWS consolidates up to six AWS services into one instance.
- In-database large language models (LLMs) support faster document analysis in Amazon S3.
- New lakehouse architecture allows high-performance querying on Amazon S3 without data movement.
- Features like AutoML and generative AI support documents in 27 languages and improve machine learning model training.
- Oracle’s HeatWave is available through the Oracle Cloud Infrastructure (OCI) Always Free Service.
- Intelligent Data Lake will integrate into Oracle’s Data Intelligence Platform by 2025, offering unified governance and scalable storage.
- Oracle Analytics Cloud AI Assistant enhances analytics capabilities by using natural language queries and LLMs.
- Partnerships with Amazon Web Services and LLM providers expand Oracle’s AI ecosystem.
- The GenDev platform in Oracle Database 23ai accelerates AI application development with hybrid data management and advanced search capabilities.
Main AI News:
Oracle Corp. unveiled new AI advancements at its Oracle CloudWorld event, focusing on expanding the capabilities of its HeatWave-managed MySQL database service. New features aim to simplify AI implementation, cutting down on the need for manual integration, advanced AI expertise, and complex security setups. By utilizing HeatWave on Amazon Web Services (AWS), companies can consolidate up to six AWS services into a single instance, streamlining operations.
Key upgrades include in-database large language models (LLMs) and automation of vector store creation. Users can run LLMs on CPUs or access models from Amazon Bedrock, enabling faster document analysis within Amazon S3. Oracle claims these features surpass those offered by Snowflake, Databricks, and Google. The new lakehouse architecture built on HeatWave allows high-performance querying of structured and unstructured data in Amazon S3 without moving the data, supporting queries across hundreds of terabytes of data.
Enhanced object storage support lets users share query results, automatically detect data changes, and run MapReduce applications. New JavaScript support in HeatWave also allows for executing stored procedures and functions, while generative AI enhancements now support documents in 27 languages. AutoML, included at no extra cost, offers advanced machine learning model training with increased data capacity and tools like topic modeling and anomaly detection.
Oracle’s HeatWave is now available through the Oracle Cloud Infrastructure (OCI) Always Free Service, offering users 50GB of storage and backup. Oracle also revealed that its Intelligent Data Lake, announced last year, will become a core part of its Data Intelligence Platform in 2025. This platform offers unified governance, scalable storage, and integration with tools like Apache Spark, Flink, and Jupyter Notebooks, allowing full-scale data lake creation and real-time data flow management. A new “zero copy” feature will allow data transfer across systems without duplication.
Oracle also introduced the Oracle Analytics Cloud AI Assistant, which translates natural language queries into actions using a built-in LLM. It can recognize Oracle Analytics workbooks and datasets while allowing integration with third-party LLMs.
Although Oracle has made significant strides in AI and cloud services, it’s still catching up with competitors. According to Gartner’s research VP Sig Nag, Oracle is advancing its multi-LLM support but hasn’t made any groundbreaking moves compared to other cloud providers. Partnerships with Amazon Web Services and LLM providers like Cohere, Llama, and Hugging Face have expanded their AI ecosystem, yet companies like Google have implemented similar strategies with Vertex AI.
Additionally, Oracle launched its AI-focused GenDev platform, combining AI technologies in Oracle Database 23ai to accelerate AI application development. GenDev’s hybrid JSON Relational Duality Views bridge relational and JSON data, while AI Vector Search allows more precise data retrieval by representing text, images, and audio as multidimensional vectors.
Oracle Database 23ai also integrates with 35 LLMs, offers GPU access for resource-intensive tasks, and includes machine learning notebooks for building deep learning models. Developers can prepare data with natural language tools or build graph models with a drag-and-drop interface. Cloud-based access starts at 39 cents per hour, with downloadable options available.
Through these developments, Oracle aims to reduce risks and accelerate the deployment of AI applications across industries.
Conclusion:
Oracle’s latest AI advancements mark a significant push to become more competitive in the cloud and AI markets. By integrating AI capabilities directly into its HeatWave service, Oracle is addressing the growing demand for faster and more scalable AI solutions. While Oracle is still catching up to competitors like Google and Amazon in AI and multi-LLM support, these new features will attract businesses looking for comprehensive, cost-effective AI services. The focus on simplifying AI development, improving data management, and expanding partnerships strengthens Oracle’s position in the enterprise cloud ecosystem, potentially driving increased adoption of its cloud infrastructure and AI offerings. This move could reshape Oracle’s market presence by offering a more accessible and streamlined AI platform.
