Guide for Using Knowledge Base in Workforce (Data/Documents Uploads)

1. What is Knowledge Base in Workforce?

Knowledge Base in Workforce refers to the documents, images, spreadsheets, and other data types that users can upload into the system. This data is stored in a vector store, allowing Teammate to perform in-depth analysis, execute vector searches, and derive insights when interacting with users. Knowledge uploads are essential to enabling Teammate to understand and utilize the content to provide relevant responses and actions.

2. How to Use the Knowledge Base Section in Teammate Builder

To utilize Knowledge in Workforce you can add documents and data through the Knowledge Base Section of the Teammate Builder. This is a key step in providing your AI Agent with the context it needs to make informed decisions and respond accurately. Below is a detailed guide on how to use this section effectively:

Navigate to the Knowledge Base Section :

Access the Teammate Builder in the Workforce interface.

Go to the Knowledge Base Section where you can add documents and data that you want your AI Agent to use.

Add Document Collection Tool

Upload Your Document :

Click the Upload File(s) button to select the file you wish to add. Supported file types include text documents (e.g., Word, PDFs), spreadsheets, presentations, and images.

Choose among available options for analyzing documents

Enable toggle called 'Enable text analysis' to use vector search with the uploaded document

Enable toggle called 'Enable code analysis' to use Code Interpreter tool with the uploaded document

Enable toggle called 'Enable image analysis' to use AI Vision with the uploaded document. It is available only if the uploaded file is in one of supported image formats

If Text analysis is enabled provide a Description:

Expand Show Additional Settings section under the tool

Replace a default Description that briefly explains the purpose and content of the document. This description helps the AI Agent understand when it should use this particular piece of Knowledge during function calls. Description is only applicable when text analysis is enabled.

Configure other Advanced Settings :

Chunking : Specify whether the document should be divided into smaller chunks. Chunking helps in processing larger documents more efficiently, allowing the AI Agent to better handle detailed queries.

Chunk Overlap : If chunking is enabled, you can define the chunk overlap . Overlapping chunks ensure that context is not lost between sections of the document, which helps in maintaining coherence during analysis.

Embedding Model Selection : Choose the model that will be used for embeddings. The embedding model determines how the content of your document is converted into vector representations for effective searching and analysis.

Additional Details : You may specify additional configurations, such as tagging specific sections or indicating sensitive content, which will help the agent make informed decisions regarding access and processing.

Save and Index :

Once you've uploaded the document and specified all necessary details, click Save . The document will be processed and added to the vector store, making it readily available for analysis and interactions.

3. How Can I Use Knowledge to Enhance My AI Workflows?

Knowledge uploads are fundamental to empowering Teammate with contextual information. For example, you can upload contracts, reports, product manuals, or spreadsheets, and Teammate will use these documents to answer questions, summarize information, and perform analysis. This feature enhances your AI workflows by enabling the agents to access rich information in real-time, which improves their accuracy and functionality.

4. What Types of Knowledge Can I Upload to Teammate?

Workforce supports a wide variety of data types that can be uploaded as Knowledge, including:

Text Documents : Such as Word documents, PDFs, and text files.

Spreadsheets : Excel files or CSV files containing structured data.

Images : Photographs, diagrams, or scanned documents for visual data processing.

Presentations : Slide decks that can be used for content extraction and summarization.

These data types are processed and indexed in the vector store to facilitate quick retrieval and understanding.

5. Benefits of Using Knowledge in Workforce

Using Knowledge comes with several key benefits:

Enhanced Contextual Understanding : Uploading documents allows teammates to understand the specific context of your business.

Improved Accuracy : Teammates can refer to precise details from uploaded documents to provide accurate responses.

Scalability : You can upload a large volume of documents, enabling the agent to have access to a comprehensive information base.

Low-Code Integration : Simply upload files without needing any special formatting or configuration, making the process accessible to non-developers.

6. Pre-Built Capabilities for Handling Common Knowledge Uploads

Workforce come with a wide range of Tools that help handle common Knowledge uploads such as:

PPT Integration : Summarize or extract insights from presentations.

Word Doc Analysis : Understand and analyze textual information in Word documents.

These pre-built capabilities allow you to get started immediately with analyzing uploaded content.

7. Can Multiple Knowledge Files Be Used Within a Single Teammate?

Yes, Workforce are designed to be highly modular, allowing you to use multiple Knowledge files within a single workflow. For instance, you could upload a set of technical manuals and financial spreadsheets , and the Teammate would be able to answer questions or perform analysis across both sets of documents seamlessly.

8. Where Can I Find More Information or Support for Knowledge in Workforce?

For more detailed information on using and configuring Knowledge in Workforce, you can visit the TorqCloud Documentation Center or join our Community Forum to get insights from other users. Our support team is also available to help you with specific questions related to your AI workflow and Knowledge configurations.

Updated on: 06/02/2025