Menu

Platform foundation

Hospitality knowledge, business context and AI.

Loop uses AI services, including OpenAI, to understand requests, prepare recommendations and support authorized work.

Current scope

Feature-dependent processing

The context sent to an AI service depends on the feature and the connected data available to that venue. A provider connection does not make all venue information available to every AI task.

What you need

  • Authorization to use the business information and assets supplied for the requested task.
  • Manager review of recommendations and specific approval where an external action requires it.

What AI helps with

AI helps Loop understand guest and manager requests, combine hospitality knowledge with relevant business context, draft responses and campaign briefs, and propose operational actions. Recommendations can be incomplete or wrong, so your team can review and adjust the work.

What may be sent to the provider

Relevant context can include the request, conversation history, venue information, selected sales or operational data, and text or assets supplied for a creative task. Where a voice or image feature is used, the selected audio or image may also be processed. Source-specific rules, including Google review restrictions, apply before context is used.

AI is separate from permission to act

An AI suggestion does not create authority to access another venue or change an external account. Loop checks the workflow’s permissions and required approval. Automated guest replies operate within the venue’s configured service rules and the channel’s policies.

Processing and operational records

AI services process the information sent for a task under their data policies and settings. Loop’s central AI monitoring tracks usage and performance without copying full requests and replies into that log. Conversation and work records have separate retention rules described in the Privacy Policy.

Technical details
  • The current AI control layer uses OpenAI API services with task-specific model selection.
  • Central usage monitoring records model, token usage, timing, status and hashed request references rather than raw prompt or response text.
  • Provider-side retention and feature-specific data handling differ from Loop’s own application records. See Security & AI and the Privacy Policy for the relevant boundaries.

Questions about your venue’s setup? Contact LoopdUp.