Oracle Generative AI: New services for enterprises

Oracle Generative AI: New services for enterprises

Oracle has announced new services for enterprises that want to leverage generative AI models in their workflows and applications. The services include updates to its OCI Generative AI service and the introduction of OCI Generative AI Agents and Data Science AI Quick Actions.

How OCI Generative AI service works and what it offers

OCI (Oracle Cloud Infrastructure) Generative AI service is a managed service that allows enterprises to access and integrate large language models (LLMs) into different use cases via API. The service was first introduced in September and is now generally available.

The service provides access to Cohere LLMs and Llama 2 70 billion parameter models. The Cohere LLMs support more than 100 languages and can generate natural language text from any prompt. The Llama 2 models are expected to be added in the coming weeks, according to Oracle.

The service enables enterprises to use LLMs for various purposes, such as content creation, summarization, translation, question answering, and more.

How OCI Generative AI Agents works and what it can do

OCI Generative AI Agents is a new service that enables enterprises to pull information from their data sources using natural language. The service is currently in beta, with the retrieval augmented generation agent as the only agent available.

The service is built on OCI OpenSearch, a managed open-source service that lets users ingest, search, and analyze their data. With Generative AI Agents, users can either ingest data in OpenSearch or use the OpenSearch repository. The service can also connect to other data sources, such as Oracle Database 23c with AI Vector Search and MySQL HeatWave with Vector Store, in future releases.

The service also has a self-check capability that provides references and citations for the information it generates. This ensures the trustworthiness and transparency of the data.

Oracle also plans to introduce pre-built agent actions across its SaaS applications, such as Oracle Fusion, Cloud Applications Suite, Oracle NetSuite, and the Oracle Cerner electronic health records platform. These actions will allow users to interact with the applications using natural language and generative AI.

How Data Science AI Quick Actions works and what it can do

Data Science AI Quick Actions is another new service that gives users access to a selection of open-source LLMs from Meta and Mistral AI. The service will launch in beta next month.

The service allows users to quickly experiment with different LLMs and use them for various tasks, such as text generation, summarization, paraphrasing, and more. The service also provides metrics and feedback to help users evaluate the performance and quality of the LLMs.

How Oracle Generative AI differs from other AI services

Oracle Generative AI is a suite of services that aims to make generative AI models applicable and useful for enterprises. Vinod Mamtani, the vice president of Generative AI Services at Oracle, said that the focus is on solving enterprise use cases with generative AI.

During an interview, Mamtani emphasized the importance of GenAI models’ capabilities and their focus on solving enterprise use cases.

However, some of the services offered by Oracle Generative AI are similar to those offered by other AI vendors. For example, OCI Generative AI service is comparable to AWS Bedrock, which also offers LLMs from Cohere and Llama via API.

Gartner analyst Sid Nag mentioned that the focus is on keeping up with other hyperscalers. This includes using and integrating GenAI in its range of applications, which will enhance the intelligence of these applications.

How Oracle Generative AI stands out and what to expect

What makes Oracle Generative AI stand out is its investment in infusing generative AI technology into its SaaS applications, said Ventana Research analyst David Menninger.

Menninger stated that the initial impact of GenAI will occur in various applications.

By integrating generative AI into its applications, such as Oracle Fusion, Oracle is providing more value to its customers, who can use generative AI to enhance their productivity and experience.

Menninger stated that organizations will derive greater immediate value from the extensions they have implemented in the Fusion applications compared to developing a custom GenAI capability.

While custom applications are important, Oracle is not the only vendor that provides them. For example, Microsoft also offers generative AI tools for developers and enterprises.

However, by offering generative AI services for applications, data, and use cases, Oracle is covering a wide range of needs and scenarios for its customers.

Oracle Generative AI is still in its early stages and has room for improvement. For example, OCI Generative AI Agents only has one agent available, and Data Science AI Quick Actions is not yet launched.

Menninger stated that additional agents would be beneficial in a similar manner to how these application capabilities are beneficial.

Oracle’s gradual roll-out approach shows that cloud providers face challenges and trade-offs in delivering generative AI services in a fast-paced market.

Menninger said that if they had unlimited resources, they would be eager to release everything at once. However, he acknowledged that this is never a realistic possibility.

Oracle Generative AI is a suite of services that enables enterprises to integrate and use generative AI models in various use cases and applications.