Today, we’re announcing Oracle Cloud Infrastructure (OCI) Generative AI Agents with retrieval-augmented generation (RAG), which includes support for Oracle Database 23ai. Building on our previous beta announcement, this latest generally available release represents a significant leap forward in our journey to empower developers to easily build conversational agents with RAG capabilities, and with Oracle Database 23ai support, OCI Generative AI Agents brings AI to your data. RAG combines the power of large language models (LLMs) with enterprise knowledge repositories to help ground model prompt-responses in business data. OCI Generative AI RAG Agent is a ready-to-use solution for users to converse with documents in their knowledge bases. 

The OCI Generative AI RAG Agent will be generally available on September 25, 2024.

Key Features

Support for more data sources: Oracle Database 23ai AI Vector Search

We’ve added support for Oracle Database 23ai, combining the power of OCI Generative AI Agents with your database vector search. Oracle Database 23ai includes advanced vector search capabilities, enabling efficient handling of unstructured data like text and images by converting them into vector representations. As a result, you can get more accurate and relevant search outcomes, even for complex queries. The feature is designed for high performance and scalability, providing quick search responses across large datasets. Fully integrated into the Oracle ecosystem, it also supports advanced analytics, allowing you to combine and analyze structured and unstructured data seamlessly. Customizable search parameters further enhance precision, better tailoring the experience to your specific needs.

The first of a series, OCI Generative AI Agents provides ready-to-use RAG capabilities, enabling customers subscribed to Oracle Database 23ai services to get started without spending cycles on manual processes, such as agent planning, retrieval, reranking, generation, and integration, and provides a self-check to help reduce hallucinations.

Streamlined agent setup

Our new RAG Agent features a simplified setup process, designed to provide you with a more intuitive user experience. With fewer steps and an improved user interface, setting up the agent is now quicker and more straightforward, allowing you to get started with minimal effort, and enabling users with limited technical expertise to deploy and benefit from RAG swiftly.

Enhanced response quality and relevance

Using state-of-the-art LLMs, the RAG Agent offers improved response quality and relevance. This enhancement helps ensure that your conversations with the AI are more accurate and helpful, making interactions more valuable and insightful. Whether you’re seeking quick answers or in-depth explanations, the RAG Agent delivers superior conversational experiences.

Advanced data retrieval and reranking capabilities

The new RAG Agent includes cutting-edge data retrieval and reranking functionalities. These capabilities allow the agent to sift through vast amounts of data and present the most relevant information, enhancing the depth and precision of insights you receive. This feature is particularly beneficial for queries that require detailed and nuanced responses.

Hybrid search

The new hybrid search feature combines traditional keyword search with advanced semantic search capabilities. This hybrid approach helps ensure that you receive the most relevant results, whether you’re searching for structured or unstructured data, enhancing the overall quality of your information retrieval processes.

Managed knowledge bases with integrated vector store and data ingestion pipeline

We’re also introducing knowledge bases, a feature that includes an integrated vector store and a robust data ingestion pipeline specifically designed for unstructured data. This feature is perfect for customers who need a managed solution, enabling efficient storage and retrieval of complex data. The integrated vector store allows for sophisticated data indexing and retrieval, making your information management more efficient.

Simplified support for BYO knowledge bases: OCI Search with OpenSearch

Our RAG Agent continues to support Bring Your Own (BYO) OCI Search with OpenSearch, and we’ve streamlined the setup process for even greater ease of use by providing greater flexibility and control over your search configurations. You can now customize and optimize search functionalities according to your specific needs. This improved setup experience makes integrating your own OpenSearch instance more straightforward than ever before.

Enhanced content moderation

To help maintain the integrity and appropriateness of the content, we’ve introduced a robust content moderation feature. This tool automatically filters and flags potentially harmful or inappropriate content, contributing to a safer and more secure AI interaction.

Multimodality

Expanding the versatility of the RAG Agent, support for multimodality allows the agent to process and integrate data from multiple sources and formats, including text and images, such as charts and graphs. With this enhancement, the agent provides more comprehensive and contextually rich responses, catering to a wider range of use cases.

You can use the OCI Generative AI RAG Agent for a variety of use cases, including:

  • Call center optimization: Operators increase customer satisfaction through more accurate responses and a higher volume of query resolution.
  • Revenue intelligence: Finance teams understand customer purchase history and trends by asking natural language questions instead of running reports.
  • Knowledge management: Business users can sift through large data stores to find relevant information to help increase efficiency and productivity.

 

Why Choose the OCI Generative AI RAG Agent?

OCI Generative AI Agents is designed to enable you to solve business problems with smarter, more efficient AI solutions. With the general availability of the OCI Generative AI Agents RAG agent, we have continued to focus on the following areas:

  • Seamless integration: Easily integrate and deploy the agent with your existing systems and workflows.
  • Scalable performance: Built on OCI’s robust infrastructure, the agent provides high scalability and performance to keep up with your growing data and interaction needs.
  • Improved accuracy: State-of-the-art LLMs and data processing technologies provide more accurate and relevant AI responses, improving your decision-making processes.
  • Integration with the existing Oracle portfolio: Oracle Database 23ai AI Vector Search is fully integrated with the broader Oracle portfolio, allowing users to combine the OCI Generative AI RAG Agent with other Oracle tools and services. This integration simplifies deployment and enhances the overall value of using Oracle’s cloud and database technologies.

Get started

The OCI Generative AI RAG Agent is now available in the US, with other regions planned to follow. To get started, visit our documentation or contact your Oracle representative.

Stay tuned for more upcoming feature announcements as we continue to innovate and enhance Oracle Cloud Infrastructure Generative AI Agents. We’re committed to delivering a cutting-edge AI experience that meets your evolving needs. We plan to add capabilities and tools to support conversing with structured data and unstructured data and supporting automated end-to-end workflows with any tools through a natural language interface.

For more information see the following resources:

 

Future Product Disclaimer
The preceding is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, timing, and pricing of any features or functionality described for Oracle’s products may change and remains at the sole discretion of Oracle Corporation.