Oracle has launched the Oracle Cloud Infrastructure (OCI) Generative AI service, marking a leap in the integration of generative AI technologies. This fully managed service seamlessly incorporates large language models (LLMs) from Cohere and Meta Llama 2, catering to diverse business use cases.
"Oracle's AI focus is on solving real-world business use cases to enable widespread adoption in the enterprise. To do this, we are embedding AI across all layers of the technology stack by integrating generative AI into our applications and converged database, and offering new LLMs and managed services—all supported by a fast and cost-effective AI infrastructure. Instead of providing a tool kit that requires assembling, we are offering a powerful suite of pre-built generative AI services and features that work together to help customers solve business problems smarter and faster," said Greg Pavlik, senior vice president, AI and Data Management, Oracle Cloud Infrastructure.
In June last year, Oracle collaborated with Cohere to integrate generative AI services into Oracle Cloud Infrastructure (OCI).
IBM has also integrated Meta's Llama 2 into its watsonx platform.
The OCI Generative AI service now boasts multilingual support for over 100 languages, elevating its accessibility on a global scale. In addition, users can benefit from an enhanced GPU cluster management experience, ensuring optimal performance for their generative AI applications.
Acknowledging the importance of customization, Oracle has introduced flexible fine-tuning options, allowing customers to tailor generative AI models to their specific needs. This feature provides adaptability and precision in addressing text generation, summarization, and semantic similarity tasks.
The company emphasizes simplicity in integrating generative AI into the technology stack, offering a managed service accessible through API calls. With a focus on data security and governance, customers can seamlessly embed generative AI into their systems while ensuring robust protection of sensitive information.
Retrieval Augmented Generation (RAG)
In the beta release, the OCI Generative AI Agents service introduces RAG, enabling customers to refine models using their own data. This combination of LLMs and enterprise search through OCI OpenSearch delivers contextualized results, facilitating natural language interactions with diverse enterprise data sources without the need for specialized skills.
Integration across its suite of cloud applications
Moreover, Oracle extends the reach of generative AI by integrating it across its suite of cloud applications, covering ERP, HCM, SCM, and CX. This strategic integration allows customers to leverage the latest generative AI innovations seamlessly within their existing business processes. With Autonomous Database Select AI, customers can merge private enterprise data with the creativity of generative AI, accelerating application development and fostering innovation.
In addition, Oracle is committed to supporting the broader AI ecosystem by expanding the capabilities of OCI Data Science. The upcoming OCI Data Science AI Quick Actions feature, set to be in beta next month, enables no-code access to various open-source LLMs, including renowned providers like Meta or Mistral AI.
Prior to these enhancements, Oracle unveiled the Fusion Data Intelligence Platform, a data, analytics, and AI platform designed to enhance business outcomes for Oracle Fusion Cloud Applications users.
