What's the best way to keep a business knowledge base up to date in Flowise?

[oliviacrter]

Hi everyone,

I'm building a Flowise assistant for a real service-based business and one challenge I'm running into is keeping the knowledge base synchronized as content changes.

The business website includes:

Service pages
FAQs
Pricing information
Blog articles
Location-specific content

I'm trying to decide between a few approaches:

Re-index the entire knowledge base on a schedule.
Update only pages that have changed.
Separate FAQs from long-form articles.
Keep structured business data outside the vector store and retrieve it differently.

For those running Flowise in production:

How do you keep business content updated?
Do you separate static and frequently changing content?
Have you found a workflow that scales well as the website grows?

I'd appreciate hearing about real-world implementations.

[oliviacrter]

Small update after a few more days of testing.

I've moved away from treating the entire website as one collection and started separating service pages, FAQs, blog articles, and location pages into different document groups. Retrieval quality has been much more consistent for customer support questions, especially when the assistant needs to distinguish between evergreen information and content that changes more frequently.

For anyone interested, the website I'm testing with is a real local service business rather than a synthetic dataset, so it includes service pages, FAQs, pricing information, location-specific content, and long-form guides. Using real business content has exposed edge cases that never appeared when I was testing with sample documents.

My next step is comparing incremental updates against scheduled full re-indexing to see which approach scales better as the site grows.

If anyone has implemented automatic change detection or selective re-indexing in Flowise, I'd still love to hear how you approached it in production.