Guest Intel is an AI feature that reads the entire guest profile and produces a concise, Swedish-language summary in three sentences: who the guest is, what they like, and how staff can deliver extra-great service. Think of it as an electronic "briefing" before arrival.
The purpose
When a VIP or regular walks in, you want staff to instantly understand who is in front of them. Traditionally that requires someone to read through order history, bookings, tags, and notes. It takes 2-3 minutes and is impractical in the middle of a service rush. Guest Intel does the job in one second with three sentences.
How it works
Behind the scenes:
Cache 24 hours – the right balance between "always current" and "cost-efficient". A guest rarely changes in 24 hours, so fast retrieval.
Example summary
Guest: "Anna Eriksson", 15 visits, 8,200 kr total, last visit 5 days ago, favorite = gravlax, tag = "gluten allergy".
"Anna is a regular who visits at least twice a month and loves red fish, especially our gravlax dish. She is careful with her gluten allergy – serve her dishes without the bread basket. Tip: She appreciates personal attention and usually orders a glass of Chablis."
Three sentences. Max 60 seconds to read. No fluff phrases.
When should you run Guest Intel?
Requirement: minimum 2 visits
Guest Intel does not generate a summary if the guest has fewer than 2 visits. Why? A single visit does not provide enough basis for meaningful insights. Instead the message appears: "Guest needs at least 2 visits for Guest Intel."
Manual refresh
If you just added an important note and want the AI to read it immediately:
Be careful about spamming this button – each regeneration costs AI tokens. Normally the 24-hour cache is enough.
Security and GDPR
Waiter Intel – the same feature in POS
In the POS, servers have the same feature under the name Waiter Intel. When a guest is checked in at a table, the AI summary appears directly in the POS sidebar. No extra clicks. The same cache is shared between admin and POS – if Boss runs a refresh in admin, it is visible in POS instantly.
Read more in the separate article "Waiter Intel – AI Guest Context for Staff".
FAQ
Q: Can the AI be wrong? A: Yes, sometimes. The AI generates based on data – if the data is incomplete, the summary is incomplete. Treat it as a hint, not absolute truth. Staff should always verify allergies and special requests directly with the guest.
Q: Is the summary printed on receipts? A: No. It is an internal staff tool, never visible to the guest.
Q: Can the guest see their own summary? A: Not today. In the future (Q3 2026) we plan a guest portal where guests can see what you know about them – part of the GDPR initiative for increased transparency.
Q: Is the summary in English for foreign guests? A: The language is set by the restaurant's locale setting. Swedish restaurant = Swedish summaries. If you run a restaurant in Sweden but want English summaries – contact support.
Q: What happens if the cache is corrupted? A: Click "Refresh summary" – that forces a new generation regardless of cache status.
This feature is part of Vendion Marketing.
Curious how it looks in practice? Read more about the product or book a short demo.
Was this article helpful?