With the introduction of multiple knowledge-base source types (web crawlers, FAQs, tables, rich text, file uploads, and more), the Knowledge Base becomes sluggish very quickly. As more sources are added, the AI has to sift through an increasingly large dataset which slows down response time and increases token cost for every answer.

A simple but powerful improvement would be the ability to name each source or (even better) assign tags to each source.

Why this matters

Many businesses using HighLevel have multiple facets, such as restaurants, event venues, wineries, medical offices, multifamily services, etc. Right now, every AI request must search all sources even if the intent is extremely narrow.

Example

If I have a business with restaurant info, event listings, and room rental details, I might have:

• a web crawler pulling Eventbrite pages

• a file upload containing rental policies

• a FAQ table for restaurant questions

If I could tag the Eventbrite crawler with “events,” then in my Voice AI or Conversation AI prompt I could clearly instruct:

“If the caller or user is asking about upcoming events, only reference sources tagged ‘events’.”

Benefits

• Faster response time

• More accurate answers

• Lower token consumption

• Cleaner prompt design

• Easier management for multi-department businesses

Request

Please add the ability to name or tag each Knowledge Base source and allow AI prompts to call tags directly. This would dramatically improve performance, cost efficiency, and answer quality for Voice AI and Conversation AI.