Скачать или смотреть EPM Digital Assistant for Planning

Users can leverage Oracle Digital Assistant (ODA) and Generative AI (GenAI) capabilities within Oracle Enterprise Performance Management (EPM) modules to significantly enhance data discovery and reporting narration through natural language interaction and automated processes.
Here is how users can leverage these features:
1. Leveraging Oracle Digital Assistant (ODA) for Data Discovery and Actions
The Oracle Digital Assistant (referred to as the EPM Digital Assistant within EPM) functions as a chatbot that enables users to interact with the system using natural language, effectively serving as an intelligent agent.
A. Natural Language Interaction and Querying:
• Navigation Bypass: ODA eliminates the need for users to follow complex navigation paths to perform data discovery. Users can get answers and data on their fingertips.
• Data Retrieval: Users can query for data using simple prompts, such as "get data for IFS".
• Dynamic Data Discovery: ODA allows users to include multiple data items in their query, such as asking to see "service revenue" alongside "other revenues" for the same intersection.
• POV Modification: Users can easily change the Point of View (POV)—including markets, accounts, or scenarios (e.g., changing from predictive forecast to plan)—by simply writing a statement within the chatbot.
• Data Visualization and Pivoting: Users can request to pivot the data (e.g., putting revenue on top) and define what they want in the rows and columns.
• Accessibility and Channels: ODA supports both voice and text commands, reducing training time. It can be integrated with external channels like Teams, Zoom, or Slack, allowing users to query data (e.g., net margin for a specific period) without needing to log into the planning application itself.
B. Executing Actions and Workflow:
• Running Business Rules: ODA is not restricted to data querying; users can perform actions, such as running business rules (e.g., "run rules for aggregate product") directly through the chat interface.
• Monitoring Jobs: Users can view the status of jobs that have been run in the application (such as data exports or rule executions).
• Task Management: If configured, ODA can show users tasks that are "at risk" and allow them to download reports or have them sent via email.
• Documentation Access: Users can access standard EPM documentation by querying the assistant (e.g., asking "what is planning").
• Customization: While the system comes with out-of-box skills, users can build custom skills or leverage the PaaS platform to add custom prompts, visualize data in graphs, and enable downloads.
2. Leveraging GenAI for Enhanced Narration and Reporting
GenAI features are embedded in EPM modules, particularly in narrative reporting, to provide enhanced summarization and automated commentary.
A. Automatic and Dynamic Narration Generation:
• Report Summarization: GenAI enhances report summarization, providing a common summary of all reports and generating narrative reports that highlight variances, bias, or comparison factors.
• Touchless Operations: The narration features support "touchless operations" where the narration automatically changes if the user alters the POV or members being viewed.
• Types of Analysis: The narrative reporting features offer three patterns for generation: describing exceptions, comparative analysis, and causality for exceptions.
B. Customization and Control over Narratives:
• Overriding Text: While system-generated text for insights and commentary may initially be fixed, within the narrative reporting feature, users have control to override the generated text. The system-generated output acts as a suggestion that users can alter based on their preferred wordings.
• ML File Modification: Users can alter the machine learning (ML) file associated with the narration features to insert their own set of narrations, ensuring the commentary reflects the organization's style (e.g., "the Sidra way of style of writing").
• Disclaimers: Users can add disclaimers, such as noting that the text is machine-generated.
C. Security and Deployment:
• Embedded Feature: The GenAI features for narration are embedded into the instance itself and do not connect externally to the GenAI feature, helping to keep data secure.
• Cost Efficiency: Since the GenAI capability is embedded, there is no extra cost required for text generation, regardless of the volume of text generated.
• Uniform Summarization: GenAI promotes uniform summarization across reports.