Oracle: AI in the Technology Stack: AI in Modern Business Processes

Oracle: AI in the technology stack

January 24, 2024 at 10:15 am | Jörg Schroeber

Oracle has announced the general availability of its Generative AI service for its cloud infrastructure. This aims to enable companies to benefit from advances in generative AI.

OCI Geneative AI is a fully managed service that integrates large language models (LLMs) from Cohere and Meta Llama 2. According to Oracle, companies can use it to address a variety of use cases: OCI Geneative AI supports more than 100 languages and provides improved GPU cluster management and tuning options Flexible. The new service can be used in Oracle Cloud and on-premises via a dedicated OCI region.

“When it comes to AI, Oracle is focused on solving real business cases to enable broad adoption within the company. To achieve this, we have integrated AI into all layers of the technology stack. To do this, we integrate generative AI into our applications and converged databases and offer courses “This is powered by a fast, cost-effective AI architecture,” said Greg Pavlik, senior vice president of AI and data management at Oracle Cloud Infrastructure. “Rather than providing a toolkit with individual components, it provides a powerful of pre-built generative AI services and functions. This helps customers solve their business challenges smarter and faster.

The Cohere and Meta Llama 2 language models are available as a managed service that can be consumed via an Application Programming Interface (API) request. Users can use it to solve problems in areas such as text writing, summaries, and semantic comparisons. They can also easily and securely integrate generative AI into their technology infrastructure, while ensuring end-to-end data security and management, Oracle said. Models can be improved using Retrieval Augmented Retrieval (RAG) techniques with your own data in order to better understand internal processes.

The OCI Geneative AI Agents Service, currently in beta, uses a RAG agent to connect LLM capabilities with institutional search based on OCI Open Search. This enables the delivery of contextual results enriched with company data. The agent also allows users to communicate with various company data sources in natural language. No special knowledge is needed for this. Even with dynamic data storage, the information retrieved is current and includes a reference to the original data source.