Oracle releases cloud AI service to assist enterprises in implementing and optimizing language models

Database giant Oracle recently released the Oracle Cloud Infrastructure Generative AI service and introduced a series of innovative technologies to provide large enterprises with opportunities to take advantage of the latest generative AI technology achievements.

OCI Generative AI Service is a fully managed, cloud-hosted offering that leverages powerful large-scale language models such as Meta Platform’s open source Llama 2 and Cohere’s proprietary algorithms and integrates with existing systems . In this way, enterprises can automate many manual business processes that are currently performed by humans. OCI Generative AI services help businesses improve efficiency and productivity and reduce error rates by providing automated solutions. This service not only reduces the need for human resources, but also provides faster, accurate and consistent results. By integrating with existing systems, OCI Generative AI services can seamlessly interact with an enterprise's workflow and data, enabling more efficient business processing.

According to Oracle, the OCI Generative AI service supports more than 100 languages, provides an improved cluster management experience for GPUs, and flexible fine-tuning options.

Oracle said their new service can call various models through application programming interfaces to solve tasks such as text generation, summarization and semantic similarity. Customers can securely embed generative AI into their existing technology stack, and the service adheres to strict data security and governance protocols.

In June last year, Oracle released the OCI Generative AI service, which is billed as a replacement for Microsoft's Azure OpenAI service. This service allows customers to consume large language models and interact with them in a conversational manner. This model is able to respond based on the content of human cues.

While Cohere’s AI models and Llama 2 are widely regarded as very powerful, they may not be that practical for most enterprises unless internal data can be leveraged to enhance and refine their capabilities. In order to solve this problem, the OCI Generative AI Agents service came into being. This service allows customers to use their own data to enhance the capabilities of large language models, making them more suitable for the specific needs and scenarios of the enterprise. Through the OCI Generative AI Agents service, enterprises can independently customize and train language models to adapt to their specific business needs and obtain higher value and benefits from them. This enhanced approach based on internal data ensures personalization and customization of models, allowing businesses to better meet customer needs and provide better services.

OCI Generative AI Agents are currently in Beta testing. It uses retrieval-augmented generation technology to fine-tune generative AI models to provide more useful, contextual responses. The service provides access to RAG agents, which leverage enterprise search capabilities built with OCI Open Search to help large language models obtain the latest information to provide more accurate responses and predictions. In this way, OCI Generative AI Agents can provide highly personalized and real-time answers to help users better solve problems and meet needs.

While the initial release supports RAG via OCI OpenSearch, Oracle plans to further integrate more data search and aggregation tools in the future. These include Oracle Database 23c with AI Vector Search and MySQL HeatWave with Vector Store. In addition, Oracle will support pre-built agent operations within Oracle's suite of software-as-a-service applications, including Oracle Fusion Cloud Applications Suite, Oracle NetSuite and Oracle Cerner. These integrations will provide users with a wider range of choices and more powerful capabilities to meet different data search and aggregation needs.

Oracle announced that the OCI Generative AI service has been launched in multiple regions of Oracle Cloud and can be accessed locally by customers through the Oracle OCI Dedicated Region platform. In addition, the service is integrated with Oracle's portfolio of cloud-based business applications, including Oracle Enterprise Resource Planning, Human Capital Management, Supply Chain Management and Customer Experiences applications. The move further enhances Oracle's database portfolio as generative AI services will be added to it through Oracle's Autonomous Database Select AI capability.

In a related update, Oracle stated that it will expand the functionality of the OCI Data Science platform and will add a new AI Quick Actions feature to the beta version next month to provide code-free access to various open source large-scale language models. Pave the way. This feature will help customers build, train, deploy and manage large language models from Hugging Face’s Transformers or the PyTorch open source library.

Greg Pavlik, senior vice president of artificial intelligence and data management at OCI, said Oracle is focused on using generative AI to solve a variety of real-world business use cases and pave the way for more enterprises to adopt this technology. "To do this, we're integrating generative AI into our applications and fusion databases, and delivering new large-scale language models and managed services that embed AI into all layers of the technology stack. We're not delivering what's needed," he said. Instead, it provides a powerful set of pre-built generative AI services and capabilities that work together to help customers solve business problems smarter and faster."

## Andy Thurai, vice president and principal analyst at #Constellation Research, believes the new service provides API-based access, which may be attractive to some enterprises because it is accessible in both cloud and on-premises environments. A bit unusual. However, the service has some limitations, notably that it only supports large language models from Cohere and Meta Llama 2, and only targets a small set of use cases, such as text generation and summarization.

Thurai said: “In terms of its overall generative AI products, Oracle is still far behind Microsoft, Google and Amazon, which all provide more powerful functions and more flexible deployment options. But , if the price is right, the option to integrate with ERP, HCM, SCM and CX applications running on OCI may make this product attractive to many enterprises. But if not, AWS may easily Win, because AWS's generative AI service is currently far ahead of Oracle."

The launch of this new service should at least help Oracle become an option for enterprises to develop generative AI. Oracle is working to become a key player in infrastructure, providing cloud-based access to Nvidia's most powerful GPUs.

The above is the detailed content of Oracle releases cloud AI service to assist enterprises in implementing and optimizing language models. For more information, please follow other related articles on the PHP Chinese website!