Hi, I have a question about implementing an AI-driven solution for a client with a frequently updated product database. The database is updated monthly, not constantly, but I need the user to ask questions about products and the bot to understand and redirect based on the product name. The challenge is that users might make typos when spelling the product names. I want to use AI to understand the user’s question, identify the intended product name, match it to the client’s database, and direct the user appropriately. If the product exists, proceed with the related flow; if not, handle it differently. Intent-based systems would work if the list was static, but that's not the case here. Feeding AI with documents in the Botpress database could work, but since products change, the AI might lack up-to-date information. Any ideas on how to solve this?

hmmm would it be possible that the KB is based on a page of website which gets updated regularly? Just an idea.

Not a bad Idea When the client publish a new product, the page list could Update. But must be hidden from SearchEngin and Web User. Any other Idea?

You can also integrate with Google Drive and, for example, Google Sheet. Any changes made to the Google documents can initiate a webhook and inform the bot about the change to index it or push it to the KB table for faster interactions. Everything depends on the type of data and required efficiency.