Oracle unveils its API-led generative AI service with Cohere’s LLMs

Just months after unveiling its three-tier generative AI strategy across multiple product offerings, Oracle has taken the covers off its new API-led generative AI service at its ongoing annual CloudWorld conference.

The new generative AI service, which is built and supported on Oracle Cloud Infrastructure (OCI) in combination with large language models (LLMs) from Cohere, is a managed service that will allow enterprises to integrate LLM-based generative AI interfaces in their applications via an API, the company said.

The API-led service is also designed in a manner that allows enterprises to refine Cohere’s LLMs using their own data to enable more accurate results.

Models, such as command, summarize, and embed, have been made available as part of the service, according to an Oracle spokesperson.

The “command” model, which is available in two sizes, generates text and can be used for text generation, text summarization, Retrieval Augmented Generation (RAG), and chat.

RAG is a new technique that combines LLMs with business data to offer more accurate results for queries.

The “summarize” model, according to Oracle, is capable of conducting abstractive summarization of the text and enables the user to configure the results with a variety of parameters to support unique use cases, such as summarizing complex documents for legal teams, technical support documentation and long email chains for busy employees.

The “embed” model, on the other hand, can translate text into numerical vectors for easy understanding of LLMs.

“Embed provides industry-leading English and multilingual models in over 100 languages for a range of use cases, including semantic search, text classification, search engine for RAG, and legacy search improvement,” Oracle said in a statement.

The generative AI service, which is currently in limited availability, will work with Oracle’s new AI Vector Search across its database offerings, such as Oracle Database 23c and its Autonomous Database, when it becomes generally available.

This service also forms the basis for generative AI capabilities embedded across Oracle’s suite of SaaS applications, including Oracle Fusion Cloud Applications Suite, Oracle NetSuite, and Oracle Cerner.

Last week, Oracle said it is betting on high demand for data, driven by generative AI-related workloads, to boost revenue in upcoming quarters as enterprises look to adopt generative AI for productivity and efficiency.

The company expects its cloud revenue to grow 29% to 31% in the second quarter of 2024, even as the overall business grows in single digits.

Other OCI-based AI updates

In addition to the new OCI generative AI service, Oracle has updated several AI-based offerings, including the Oracle Digital Assistant, OCI Language Healthcare NLP, OCI Language Document Translation, OCI Vision, OCI Speech, and OCI Data Science.

The OCI Language Healthcare NLP, according to the company, adds healthcare insights with natural language processing via new models that help the AI recognize medical terms, their relationships, and entities in health records, such as notes from prescriptions, clinical trials, and other electronic health records.

OCI Language Document Translation, too, enhances translation by adding support for new formats, such as Word, PPT, HTML, Excel, and JSON.

Oracle’s OCI Vision and OCI Speech have also been given new capabilities such as facial detection and diarization — a process to segregate audio with multiple human voices into separate segments based on the speaker’s identity.

“Diarization makes OCI Speech a valuable tool for organizing, analyzing, and extracting meaningful information from spoken interactions,” the company said in a statement. Oracle has also added a new Feature Store to OCI Data Science to manage application features developed by data science teams.

“Feature Store provides a cohesive framework where features are meticulously documented, shared, stored, and served in a streamlined manner,” Oracle said.