New improvements for generative AI
The heart of OCI’s latest innovations is the OCI Generative AI Agents service with retrieval-augmented generation (RAG) capabilities. This RAG Agent combines the power of large language models (LLMs) with private enterprise data and retrieval-augmented generation (RAG), allowing users to converse with diverse enterprise data sources through natural language, without requiring specialized skills. Users of Oracle Database cloud services and OCI Search with OpenSearch can converse with their enterprise data without complicated processes to move data and eliminate the manual processes to set up an RAG environment—instead focusing on gaining new insights and opportunities for growth.
But that’s not all. They are also introducing additional Agents that streamline insights, productivity, and efficiency, allowing users to interact with their systems in a new way. Available in alpha, SQL Agent enables natural language to SQL generation, making it easier for non-technical users to access and analyze data. Also in alpha, more agents for the Oracle Cloud Console provide natural language to advanced resource search (RQS) and natural language to billing inquiries. So, customers can use the OCI Console more efficiently than ever.
New model for OCI Generative AI
The general availability within OCI Generative AI of the Llama 3.1 models, which include the largest open source LLM to date at 405 billion parameters. This model provides customers with a powerful tool for generating human-like text and is usable in a wide range of applications, from writing assistance to summarization and analysis. Meta Llama 3.1 models are now available in sizes ranging from 405 billion parameters for AI use cases that require state-of-the-art capabilities to 70 billion parameters for more targeted workloads at a lower price point. As previously announced, OCI Generative AI continues to support the Cohere Command R, Command R+, and Embed models.
New updates for OCI Data Science
Oracle’s latest innovations don’t stop there. OCI Data Science, the AI platform for data scientists and ML engineers to develop, manage, and deploy machine learning models, released new updates to AI Quick Actions, its no-code solution to fine-tune, deploy, and evaluate LLMs. AI Quick Actions now supports the Ampere ARM A1 shape to deploy models on CPUs, and enables the import of any model from Hugging Face with a few clicks of a button, opening up the ability to easily bring thousands of open-source models into OCI.
Expanding the possibilities of AI
New capabilities are also available across languages, data formats, and industries in our OCI AI Services. For example, OCI Language’s personally identifiable information (PII) feature, which allows entities to be detected as PII, now also supports a custom rules engine. This improvement enhances overall accuracy by tailoring results to specific use cases. You can also now deploy translation services on a dedicated endpoint, offering greater control, performance, and isolation for customers who need it. Our file translation API now automatically detects document language.
We’ve also updated the translation models for 11 languages and enhanced our support for popular file formats such as SRT, JSON, HTML, and more. Lastly, as a quality-of-life improvement for developers, all our dedicated and custom model APIs will now support alias names for endpoints, reducing the need for code changes when an endpoint is updated.
The OCI Speech service synthesizes speech from text, with natural voices in text to speech and a new real-time transcription capability that includes custom vocabularies support.
The OCI Vision service can now detect labels and analyze objects, text, or faces in the entire video or within every frame in the video using pre-trained or custom models. For the healthcare industry, OCI Document Understanding now supports medical ID cards.
Real-world impact in the AI world
So, what do all these changes mean for organizations? The short answer is that Oracle’s latest innovations have the potential to drive real-world impact. With its new agents and expanded capabilities, Oracle is making AI more accessible, more efficient, and more effective with direct access to your enterprise knowledge bases. Furthermore, with its continued focus on enterprise needs, Oracle is delivering AI solutions that can integrate into daily operations, enabling organizations to gain immediate and real-world business results.
As we look to the future of AI, one thing is clear: The possibilities are endless. With Oracle Cloud Infrastructure leading the way, organizations can be confident that they have the tools and solutions they need to succeed in an increasingly AI-driven world.
Enchant Apps has a team of Oracle Cloud experts across all modules. If you are interested in learning more or have a project where you need Oracle cloud expertise, feel free to contact us!