At its annual CloudWorld conference, Oracle made a significant stride in its AI strategy, revealing the API-led generative AI service. This novel offering, powered by Oracle Cloud Infrastructure (OCI) and Cohere LLM, introduces a managed service that empowers enterprises to integrate LLM-based generative AI interfaces into their applications via an API.
Notably, this API-led service is designed with flexibility in mind, allowing enterprises to fine-tune Cohere's LLMs using their proprietary data, thus enhancing the precision of results. Oracle is providing a range of models, including "command," "summarize," and "embed," as part of this groundbreaking service.
The "command" model, available in two sizes, excels at text generation, text summarization, Retrieval Augmented Generation (RAG), and chat functionalities. RAG, a cutting-edge technique, merges LLMs with business data to deliver highly accurate query results.
Meanwhile, the "summarize" model offers abstractive summarization capabilities, enabling users to configure results with various parameters. This feature proves invaluable for complex document summarization, technical support documentation, and simplifying lengthy email chains.
The "embed" model bridges text and numerical vectors, facilitating a better understanding of LLMs. Oracle boasts industry-leading English and multilingual models across 100+ languages, catering to various use cases like semantic search, text classification, RAG search engines, and enhancing legacy searches.
While the generative AI service is currently in limited availability, it is set to work seamlessly with Oracle's AI Vector Search across database offerings, including Oracle Database 23c and Autonomous Database, upon general release. Furthermore, it is the foundation for embedding AI capabilities across Oracle's suite of SaaS applications, encompassing Oracle Fusion Cloud Applications Suite, Oracle NetSuite, and Oracle Cerner.
Oracle's strategic bet on the surging demand for data-driven AI workloads is expected to drive revenue growth in the coming quarters. The company anticipates cloud revenue to surge by 29% to 31% in the second quarter of 2024, even as the overall business grows in single digits.
In addition to the OCI generative AI service, Oracle has also unveiled enhancements across various AI-based offerings:
Oracle Digital Assistant: Updated with improved capabilities to enhance user interactions.
OCI Language Healthcare NLP: Now equipped with new models that enhance healthcare insights by recognizing medical terms, their relationships, and entities in health records.
OCI Language Document Translation: Expands its translation capabilities to support new formats, including Word, PPT, HTML, Excel, and JSON.
OCI Vision and OCI Speech: These offerings now feature advanced capabilities such as facial detection and diarization, which segregates audio with multiple human voices into separate segments based on the speaker's identity.
OCI Data Science: Oracle has introduced a Feature Store to manage application features developed by data science teams, streamlining the documentation, sharing, storage, and serving of features.
Oracle's commitment to advancing AI and generative capabilities underscores its dedication to meeting the evolving needs of enterprises in the era of AI-driven productivity and efficiency.