What is QuickRead?
QuickRead is an AI capability powered by Whatfix. It provides summarized answers to your end users' search queries within the Self Help widget.
QuickRead uses information from your Whatfix content and the crawled content from your integrated repositories to generate a response to your end users' queries.
QuickRead provides responses with links, called Sources, that direct you to the source of information, wherever possible.
What are the benefits of QuickRead?
QuickRead offers the following benefits:
For end users:
Significantly reduces search time.
Improves the productivity of end users.
Provides faster contextual help for the end user.
Reduces the cognitive load of end users by providing concise and relevant information.
For content creators:
Enables content creators to improve the quality of their help content based on end users' feedback on the generated responses.
Enhances the end-user adoption of the content created by the content creators.
How does QuickRead work?
QuickRead uses Large Language Models (LLMs) on Whatfix content and aggregated content to generate responses to end users' queries. Once enabled for an account and pushed into Production, QuickRead provides responses for any end-user query that contains more than one word, provided relevant content is available.
When you enter your query in the Explorer, QuickRead automatically searches for a question based on the content in the Self Help segment or its search scope. QuickRead then provides a response to that question. The question varies for each end user, as it is chosen based on the user's role.
QuickRead can respond to queries with information from the following Whatfix content types:
Repository links
Articles
Flows
PDFs
Note:
For QuickRead to generate responses based on PDF content, the PDF file size must be less than or equal to 200 MB for your account. This means you can upload multiple PDFs, but the combined total size of all uploaded PDFs should not exceed 200 MB. This limit can be increased up to 500 MB. To increase the limit, contact support@whatfix.com.
Once QuickRead is enabled, all Self Help segments on your application have QuickRead. QuickRead generates its responses using the information within the content in a Self Help segment's search scope. It further filters information based on the role tags attached to your content. Information within content that contains role tags is not shown to end users who do not have the correct roles.
QuickRead Support in Repository Links
QuickRead reads two categories of content from the integrated repository:
Unstructured content as text only and answers questions based on meaning. It includes PDFs, Word files, presentations, web pages, and knowledge base articles.
Structured content organized in rows and columns. This structure supports a more precise set of queries. QuickRead reads structured spreadsheet content within your integrated repositories and includes CSV, XLS, and XLSX formats.
Info:
QuickRead supports structured spreadsheet content only within repository links. You cannot upload a CSV or Excel file to the Dashboard as a Whatfix content type.
Supported query types
QuickRead supports the following query types for spreadsheet content from the integrated repository:
Query type | Description | Example query |
|---|---|---|
Text lookup | Finds a record by name or description. | Provide details of Brooklyn Center |
Numeric lookup | Finds a record by an exact ID or number, even when you enter the number on its own. | 410 |
Contextual lookup | Returns a specific attribute of a record. | Is a military center discount provided to center 450? |
Filtering | Returns records that match a condition. | How many centers are there around New York? |
Multi-condition filtering | Returns records that match more than one condition. | List all lease centers in California |
Aggregation | Counts or summarizes values across rows. | How many centers have Military Waived Reg Fee enabled? |
Spreadsheet limitations
QuickRead reads a spreadsheet file with up to 2,000 columns and 10,000 rows, and the file size must not exceed 5 MB.
Note:
QuickRead reads each sheet in a workbook as a single table. A workbook can contain more than one sheet, and QuickRead reads each sheet as one table.
QuickRead does not support more than one table within the same sheet.
QuickRead does not support files that go beyond the specified limits.
Guidelines and best practices
Consider the following behaviors and limitations when you use spreadsheet content with QuickRead:
QuickRead reads a spreadsheet as structured data only in the CSV, XLS, or XLSX format, within your integrated repository.
If a spreadsheet is exported to PDF, or if spreadsheet content is embedded in a PDF document, QuickRead processes it as a text document rather than as rows and columns. The query types described above do not return equivalent results from a PDF. The file must remain in CSV, XLS, or XLSX format for the spreadsheet capability to apply.
QuickRead shows the top three to five most relevant records for a broad list query, along with a link to open the full source spreadsheet.
QuickRead runs aggregation and count queries on a per-sheet basis.
QuickRead reads pre-calculated cell values directly and does not perform calculations.
QuickRead does not process charts, images, or pivot tables within a spreadsheet.
QuickRead does not support files that exceed the stated limits.
Additional Features
Info:
Expand the accordion(s) for more details.
Multi-language Support
QuickRead can understand and respond to queries in any European language supported by Whatfix. QuickRead responds in the language the query is written in, even if the source is in a different language.
Feedback
End users can provide feedback by giving a thumbs-up or a thumbs-down based on the relevance of the response.
If the end user gives a thumbs down on a QuickRead response, they have the option to comment on why they gave a thumbs-down feedback.
Analytics
Whatfix enables users to view analytics based on end-user feedback and engagement. For more information, see QuickRead Analytics.