Introduction to Oracle Analytics Cloud AI Assistant

SEPTEMBER 15, 2024 03:30 AM

Overview

Oracle Corporation began rolling out Oracle Analytics Cloud (OAC) AI Assistant (AIA) feature in late September 2024. Oracle Analytics Cloud AI Assistant is a tool that uses natural language to help users interact with data by utilizing a large language model (LLM) to understand the context of a user's question and translate it into actions within the platform. This eliminates the need to master intricate tool functions and allows users to focus on interpreting the results rather than navigating the complexities of the software itself. It acts as a conversational interface for data analysis and recognizes Oracle Analytics workbooks and datasets. AI Assistant (AIA) leverages natural language processing (NLP) and machine learning to understand complex questions and provide relevant data insights within the OAC platform.

Key Benefits

Increased Accessibility: Enabling non-technical users to analyze data effectively through natural language interactions.
Faster Insights: The Analytics Cloud AI Assistant understands the context of a user's question because it features a built-in large language model (LLM) optimized for analytics conversations and tasks that recognizes the Oracle Analytics workbook and datasets.
Enhanced Productivity: The Analytics Cloud AI Assistant streamlines workflows and enhances analyst productivity on data insights instead of mastering tools.

Core Features

Natural Language Querying (NLQ): Users can ask questions in plain language to explore data, identify trends, and gain insights without complex SQL queries.
Contextual Understanding: The AI Assistant can interpret the context of a question based on the current data set and the analysis being performed.
Data Visualization Generation: Generate relevant charts, graphs, and tables to represent the data insights.
Explanation of Results: Provide clear explanations for the insights generated, including key drivers and contributing factors.

Technical Aspects

AIA seamlessly integrates into the existing OAC interface and interacts with data models and visualizations. This underlying AI technology utilizes the large language model (LLM) to power natural language understanding and response generation. AIA includes customization options that allow you to tailor the AI Assistant to specific business needs by adjusting the underlying models or providing custom data sources.

AIA Business Use Case Examples

Manufacturing: "What products had the most returns?"
Human Resources: "What factors affect employee churn rate?"
Sales: "What is the revenue stream comparison across IT services?"

Learn more about Set-up and Deployment.

Check Availability

Index your data so it is available for Oracle Analytics AI Assistant and Home Page Ask. AIA will initially only be made available for shape sizes of 10+ OCPUs. To check if the AIA is available in your environment, display the Search tab of a dataset and confirm that you can index for the AIA.

Index a Dataset

A dataset must be indexed to make the data attributes available on the Home page and to AIA. You can index all or some of the data set's attributes and apply synonyms to make the attributes easier to search.

On the Home page or Data page, select a dataset.
Hover over a dataset, then click Actions, then Inspect.
Click the Search tab.
Click the Index Dataset For dropdown and select an option.
Click Index and select an option to index the data elements.
Select Use Recommended Index Settings to let the system determine your index settings.
Click the Languages field and select the language you want to use to produce the dataset's index.
Click Save.
Click Run Now to index your dataset immediately.
Click Refresh to check the completion status of the index.

Schedule When to Index a Dataset

You can schedule when and how often a dataset is indexed. By default, when indexing is enabled for a dataset, the dataset is indexed when it is refreshed.

On the Home page, click Navigator, and then click Data.
Click the Datasets tab.
Locate the dataset where you want to add an indexed schedule, click Actions, then click Inspect.
Click the Search tab.
In the Indexing Schedule section, click Start and then click the Select Date Time button.

Description of the illustration dataset_search_index_schedule.png
In the dropdown calendar, browse to and select a month and day. At the bottom of the dropdown calendar, click the time stamp and specify what time you want indexing to run.
Go to the Repeat every field and enter a number. Click the dropdown and select how often you want the schedule to run.
Click Save.

Index a Dataset on Demand

You do not have to wait for a dataset to be indexed after a refresh for the dataset's index scheduled to run. You can index a dataset anytime you need to make its data available in the Home Page's search results.

Make Analytics Content Easier to Search

Content authors make analytics content easier to search from the Home page by specifying synonyms for columns in datasets. For example, to make it easier for users to find data in a column named "Yield", you might specify "revenue" and "income" as synonyms. From the Home page, users can locate data in the "Yield" column by entering "revenue" or "income" as a search term.

Specify synonyms for column datasets to make analytics content easier to search from the Home page. For example, you might enable users to search on "Volume" to locate data in the column "Quantity Sold".

On the Home page or Data page, select a dataset.
Hover over a dataset, then click Actions, then Inspect.
Click Search, and make sure that the Index Dataset for Searching option is selected.

If the Index Dataset for Searching box is deselected, other options are greyed out. If you cannot select the Index Dataset for Searching, ask the Oracle Analytics administrator or dataset owner to give you read-write access.
Enter synonyms in the Synonyms field next to the column you want to update.

Description of the illustration synonyms.png
Click Save.

Add or Update a Dataset's Permissions

Oracle LiveLabs Sprint Link

You can assign users, roles, and access permissions (Full Control, Read-Write, and Read-Only) to a dataset that you create or administer.

Assign permissions to users and roles to specify who can access the dataset and what they can do with the dataset (i.e. reload data, index the dataset, or download file).

On the Home page, click Navigator and then click Data.
Click the Datasets tab.
Locate the dataset you want to add permissions to or update permissions for, click Actions, and then click Inspect.
Click the Access tab.
Optional: To modify permissions, locate a user or role and click the assigned permission you want.
Optional: To delete a user or role, hover over it and click Delete.
Optional: To add users and roles, click the Search field and type the name of the user or role you want to add. Select the user or role from the search results list to add it and click the permission you want to assign it.
Click Save.

Summary of Key steps to use the OAC AI Assistant

Enable Your Dataset: Check if your dataset is ready for use with the AI Assistant.
Index Your Dataset: Access the "Search" tab within the dataset inspector and choose to index your data for the Assistant.
Ask Questions: Once indexed, type natural language questions in the Assistant panel to receive visualizations and insights based on your data.

Putting AIA Into Use

To use the Oracle Analytics Cloud (OAC) AI Assistant, navigate to your dataset in the "Data" section, open the "Inspect" panel, go to the "Search" tab, select "Assistant", and then ask natural language questions about your data to receive visualizations and insights generated by the AI based on your dataset. Ensure your dataset is indexed for the assistant to function properly.

Tips

Access Point: You can access the AI Assistant by clicking "Auto Insights" within a workbook and selecting the "Assistant" tab.
Natural Language Queries: The AI Assistant understands natural language queries, allowing you to ask questions conversationally.
Data Interpretation: The Assistant analyzes your data and generates visualizations highlighting trends and patterns.
Customization: You can refine your queries by providing further context or specifying specific data points.

FAQ

What is Oracle Analytics Cloud AI Assistant (AIA)?
The Oracle Analytics Cloud (OAC) AI Assistant is a tool that uses natural language processing (NLP) and machine learning to help users interact with data by asking questions in plain language. It interprets the context of user queries and generates insights, visualizations, and data analyses without needing complex technical knowledge or SQL queries.
How do I enable and index a dataset for use with the Oracle Analytics AI Assistant?
To use the AI Assistant, you must first ensure your dataset is indexed. This involves selecting a dataset, navigating to the "Inspect" panel, and then using the "Search" tab to index the dataset. You can choose to index all or some of the data attributes and can set synonyms to make the dataset easier to search. Once indexed, the dataset will be available for natural language queries via the AI Assistant.
Can I customize how the AI Assistant interacts with my data?
Yes, the AI Assistant offers customization options. You can tailor it to your business needs by adjusting the underlying models or adding custom data sources. Additionally, you can refine queries by specifying more context or certain data points to improve the relevance and accuracy of the insights generated.

Conclusion

The primary value of using the Oracle Analytics Cloud (OAC) AI Assistant is its ability to simplify complex data analysis by allowing users to ask questions in natural language and generate visualizations and insights without requiring extensive technical knowledge. This increases productivity and enables faster decision-making by bridging the gap between business questions and data analysis complexity.

The critical items to focus on when deploying AIA are:

Understand the use of custom and generic LLM
Learn how to create metadata that creates "human speak"
Learn how to determine what should not be targeted to index

Artificial Intelligence requires ample set-up work to return to a reliable benefit. Most data are not ready for natural language interactions. The set-up, configuration, use of synonyms, and utilization of various large language model types are critical to the success of AIA; descriptive synonyms improve response accuracy and relevance.

SMACT Works is a technology-focused systems integrator and IT/ERP consulting firm. We deliver end-to-end consulting, managed, and implementation services for Oracle Cloud Applications, IaaS & PaaS, On-Premise PeopleSoft & EBS Applications. Headquartered in Dublin, OH, we have a global presence with North America and Asia offices. We are an Oracle Gold Partner Cloud Standard, ISO 9001, and 27001 certified delivery organization serving customers with Excellence and Integrity.