Skip to main content

Data Sources

Managing Data Sources: Fueling Your AI's Knowledge​

The "Data Sources" section is where you empower your ReadyChatAI bot with comprehensive knowledge. By providing and managing information here, you train your bot to answer customer questions accurately and efficiently, reducing the need for human intervention.

⚠️ We're in Beta! Please note that the Data Sources section, like other parts of ReadyChatAI, is currently in beta. Some features may be limited or subject to change as we continue to enhance the platform.

ReadyChatAI Data Sources Interface
The main interface for managing your AI's Data Sources, showing the Q&A tab.

Overview of Data Source Types​

ReadyChatAI offers multiple ways to feed information to your bot. You can switch between these methods using the tabs at the top of the "Data Sources" page:

  • Q and A: This is for directly adding specific question-and-answer pairs. It's ideal for common FAQs, addressing previously unanswered customer queries, and providing precise information.
  • Website: This tab will allow you to input website URLs (e.g., your company's FAQ page, product pages, or blog). The AI can then learn from the content on these pages to answer related questions.
  • Import Files: This feature will allow you to upload documents (such as PDFs, TXT files, or CSVs containing FAQs) from which the AI can extract knowledge and learn.

The primary focus of this guide will be the "Q and A" method, as it offers direct control over your bot's learned responses.

ReadyChatAI Data Sources Extended View

The "Q and A" Tab​

This tab is your central hub for manually teaching your bot. It helps you manage existing knowledge and address new queries that your bot couldn't handle.

1. Navigating Question Lists: "Unanswered Questions" vs. "All Questions"​

Within the "Q and A" tab, you can toggle between two main lists:

Unanswered Questions​

(e.g., "Unanswered Questions (9)")

This list is crucial. It shows questions that customers asked but your bot couldn't answer from its existing knowledge base. The descriptive text "These are the questions your bot needs assistance with..." highlights its importance. Prioritizing these helps you continuously improve your bot's accuracy. The table typically displays columns like:

  • Question: The exact query from the customer.
  • Products: Indicates if the question is linked to a specific product. A message like "This question will not be displayed because it is not associated with any product" may appear if product linking is required for the bot to use this Q&A effectively.
  • Last Asked: The date the question was most recently posed.
  • Channels: The social media platform (e.g., 📘 Facebook) where the question originated.
  • Status: The current status (e.g., "Open" if it still needs an answer).
  • Actions: Provides options to manage the question, such as answering or editing it.

All Questions​

(e.g., "All Questions (16)")

This view displays all the question-and-answer pairs currently stored in your bot's knowledge base. This includes questions you've proactively added and those you've answered from the "Unanswered Questions" list. The table here might show columns like:

  • Question: The stored question.
  • Products: Any product(s) associated with this Q&A.
  • Actions: Options to edit or delete the Q&A pair.
  • Answer: The stored answer for the question.

2. Answering an Unanswered Question or Adding a New One​

When you click to answer an unanswered question or use the "Add New Questions" button, you'll typically be directed to a dedicated interface:

Question Answer Interface
Adding or editing a Question and Answer pair.

The interface includes:

  • A "Back" button allows you to return to the previous list.
  • Question Asked By: May show anonymized user information if it's an existing unanswered question.
  • Note on Product Association: A reminder like "This question will not be displayed because it is not associated with any product" may appear, emphasizing the importance of linking Q&As to products where relevant.
  • Question: For new Q&As, you'll enter the question here. For unanswered ones, it will be pre-filled.
  • Answer: A text field where you provide the clear, concise, and helpful answer your bot should use.
  • Link the Question to Product(s): A crucial dropdown or selection field to associate this Q&A pair with one or more of your products. This helps the bot provide contextually relevant answers, especially if the same question might have different answers depending on the product.
  • Click the "Answer Question" (or a similar "Save Question") button to store this information and train your bot.

3. Other Key Actions on the "Q and A" Tab​

  • Add New Questions: This button allows you to proactively populate your bot's knowledge base with Q&A pairs for questions you anticipate customers might ask.
  • Export: This function lets you download your Q&A data, which can be useful for backups, offline review, or transferring to other systems.
  • Filter (🔍): The filter icon typically opens options to sort or narrow down the displayed list of questions based on criteria such as channel, status, date range, or associated product.

Utilizing "Website" and "Import Files"​

While the "Q and A" tab offers direct manual input, the "Website" and "Import Files" tabs provide powerful ways to bulk-train your bot:

  • Website Data Source: Here, you would input URLs of your relevant web pages (e.g., your company's main website, specific product pages, blog, or existing FAQ pages). ReadyChatAI would then crawl and process the content on these pages, enabling the bot to automatically learn and answer questions based on that information.
  • Import Files Data Source: This feature would allow you to upload various document types (e.g., PDF brochures, TXT service descriptions, CSV files of pre-existing question-answer lists). The AI would parse these files to expand its knowledge base, saving you from manual data entry.

Combining these different data source methods can create a highly knowledgeable and effective AI bot with minimal manual effort.

Best Practices for Managing Data Sources​

  • Prioritize Unanswered Questions: Regularly check and address questions in the "Unanswered Questions" list. This is the most direct way to improve your bot's accuracy and customer satisfaction.
  • Be Clear, Concise, and Customer-Centric: Write answers in simple language that is easy for customers to understand. Formulate questions as your customers would ask them.
  • Link to Products Accurately: When a question or answer is specific to a product or service, ensure you link it correctly. This prevents confusion and ensures contextual responses.
  • Keep Information Current: Your business evolves. If product details, pricing, policies, or services change, remember to update the relevant Q&As, website links, or imported files in your Data Sources.
  • Use a Mix of Sources: Leverage the Q&A tab for specific control, the Website source for broad knowledge from your site, and File Imports for existing documentation.
  • Test Regularly: After adding new data, use the Playground or interact with your bot on live channels to ensure it's responding as expected.