Oracle Artificial Intelligence—Platform

Machine Learning Development Environment
Fastest AI Development for Developers and Data Scientists

Accelerate AI development on any framework using best-in-class infrastructure.

Oracle’s Holistic Approach to AI

Oracle’s AI approach transforms people and organizational productivity, efficiency, and insight, economically, and at scale. Oracle embeds human-reasoning techniques in every layer of the cloud, improving automation. And it provides data scientists with the tools, frameworks, and state-of-the-art infrastructure to rapidly build, train, and test machine-learning models. The platform also leverages three unique Oracle strengths that are critical to successful model development: contextual enterprise data, advanced data and analytics management, and a high-speed cloud infrastructure.

Developers can easily integrate trained models into a wide variety of Oracle Cloud application environments, including sophisticated, conversational, AI chatbots. For analysts and data scientists, the platform’s discovery and visualization tools provide an interactive and nonprogrammatic way to support validating models. The Oracle AI platform also increases productivity for scientists, engineers, and analysts as they experiment and innovate with larger datasets.

There are key enterprise challenges that customers face:

  • Large Scale Data Management: Artificial intelligence and ML require storage and processing of vast amounts of data, along with capabilities that encompass data aggregation from a variety of sources including structured and unstructured data, data ingestion, data cleansing and normalization, and data enrichment with business-specific context. All of this is difficult to accomplish on premises.
  • Domain-Specific Knowledge: For AI and machine learning to deliver their promised value, domain- specific knowledge and data must be applied to the models. Trained, domain-specific models enhanced with firm-specific configurations and preintegrated in purpose-built SaaS applications or platform services should be provided as an option.
  • Development Complexity: Enterprises engaged in building their own ML models are confronted with complexities in setting up the environment for data ingestion, the storage and high performance compute infrastructure for training, and setting up the preferred ML frameworks and development tools. They must then ensure that the environment is regularly updated with the latest versions of tools and the latest generation of high-performance compute resources.

To address these challenges, Oracle is uniquely providing the key elements for AI enablement.

  • Large-Scale Data Management for AI: Data-management infrastructure required for ingesting large volumes of data, data cleansing and normalization, and data enrichment with meta data and context specific to the enterprise.
  • Machine Learning Development Environment: A comprehensive environment for rapid development of advanced ML models that provides quick access to a variety of ML frameworks and model development tools, a rich selection of ML algorithms, tools for collaboration among ML model developers, and APIs to easily access these AI platform services.
  • Infrastructure for AI: high-performance compute and storage infrastructure optimized for ML that combines:
    —Enterprise-class compute performance for accelerated model training
    —High performance storage combined with lowest-cost archival storage for vast training and test data
    —Highest-available network performance to enable large data transfers from various sources of data
  • Embedded AI: In addition to providing the technology foundation for developing and running AI processes and solutions, Oracle embeds AI and ML capabilities within its own business and IT services. By adding ML and cognitive interactions into traditional business and IT operational processes, users will experience greater productivity and insight.

AI Platform Cloud Service

Oracle AI Platform Cloud Service allows developers and data scientists to quickly set up a complete environment for developing ML models. It enables them to use their preferred frameworks and tools running against class-leading GPUs on a 25 Gb network to accelerate data aggregation and model training.

AI Platform Cloud Service

Conversational AI Interfaces: Intelligent Bots

Chatbots enable automated human and computer conversations using voice and text. Service and help desks are often overwhelmed with requests and can be unresponsive. Chatbots can eliminate the backlog by enabling automated responses and then handing off to human agents when necessary, thus increasing user satisfaction. Oracle Intelligent Bots include five core AI capabilities: ML, cognitive services, knowledge services, dialog and context, and data and insights. A pipeline of AI algorithms collectively works to understand and process the end user’s input expressed in freeform natural language. In addition, additional algorithms are used to support language translation, image recognition, sentiment analysis, and more. The solution provides a framework for rapid development, usability effectiveness, and ongoing learning. Oracle Intelligent Bots is part of the Oracle Mobile Cloud Enterprise platform and will be integrated into Oracle SaaS applications.

Database Analytics and Machine Learning

Oracle provides high-performance data management and ML compute instances for industry-leading price performance of AI-based applications. Using Oracle Exadata and big-data platforms, Oracle’s ML and advanced analytics capabilities deliver high-performance ML algorithms for both for both structured and unstructured data. Using these powerful data-management and advanced analytics/ML platforms, Oracle dramatically reduces the time to build and deploy enterprise predictive models and applications from weeks/months to minutes/hours.