Training - Data Sources

Data is the "backbone" of AI. Continuously providing and updating data is the key factor determining your chatbot’s response quality. Easy AI Chat supports various types of data sources:

Data from Website 

The AI Chatbot will, by default, learn from the website domain you provided. The system prioritizes important pages such as Homepage, About Us, Products, and Services.

Note: Easy AI Chat can only scan pages that are declared in the sitemap to search engines. Before loading data into Easy AI Chat, you should:

• Check your website in "Disable JavaScript" mode: Use browser developer tools to see how your site appears when JS is turned off.

• Use Server-Side Rendering (SSR): SSR ensures that your site’s content is always available to Google Bot, even with JS disabled.

• Optimize for Search Engines (SEO): Ensure your content is clearly structured and easy to understand for both users and bots.

• Use SEO audit tools: These tools can help identify SEO-related issues on your site

=> See more: Guidelines for Standardizing Data for Easy AI Chat Chatbot

Data from Sample Q&A

Sample questions serve as a supplementary data source for content not yet available on the website, directly targeting common user inquiries.
We accept text formats and hyperlinks in the answer section of sample questions.

Note: There is no limit to the number of sample questions. When creating questions, you should:

• Group questions by topic (e.g., Products, Orders, Warranty, Contact Information).

• Write questions the way customers typically ask them.

• Provide concise answers containing relevant keywords.

• Regularly update questions and answers based on customer feedback

Updating Information: You can easily edit or add new sample questions. The AI will take a few minutes to read and update the new information.

Data from Document Files

We accept the following file formats: PDF, CSV, DOC, TXT, XLS with a maximum file size of 10MB per file. The number of files allowed per account depends on the subscribed plan.

Requirements:

• PDF: Ensure that text can be extracted (not scanned images).

• CSV: Use the first row as column headers and explain the meaning of each column.

• DOC/DOCX: Use built-in styles to format text.

• TXT: Organize information with a clear structure, using spaces or special characters to separate sections.

• XLS/XLSX: Clearly name each sheet and use the first row as the header.

You can review scanned document content or rename files from the detailed document interface.

Updating Information: You can easily review scanned document content or rename files. The AI will take a few minutes to read and update new information.

Data from Single URL 

This includes links/sitemaps outside the default domain’s sitemap. These are particularly suitable for websites with multiple subdomains or businesses that own several websites but want to use a single bot to fully understand the business and provide comprehensive consultation.

Data from Single URL: Allows you to add links or sitemaps outside the default domain. This is especially useful for businesses with multiple subdomains or multiple websites that want to use a single bot for overall consultation.

Updating Link Data: You can easily click "Retrain" for individual links or for the entire dataset so that the AI can re-scan and update information at the latest time, ensuring the data stays up to date.

(Coming Soon) Data from YouTube Links