Conceptual hospitality illustration depicting an anonymous venue.
Hospitality teams work with time-sensitive information and decisions that affect a guest’s experience. AI can be evaluated against concrete tasks in that setting, such as preparing a service briefing or organising supplier information. The starting point should be a task the team can review.
The My NEO Group network identifies Salvaje Group as a partner and includes its hospitality venues. This connects the Group’s technology perspective with an operational sector. The examples below are proposed evaluation scenarios, not claims of deployed AI systems or measured results at those venues.
Begin with service preparation
An illustrative pilot could prepare a briefing from an approved event plan and a supervisor’s service notes. The draft might identify the expected sequence of service and unresolved operational questions. A manager would review it before the team uses it.
The pilot should preserve the origin of each important instruction. If an event time changes, the briefing needs to show the update and identify affected sections. A summary that reads well but carries yesterday’s timing can create additional work. Review the briefing with the people responsible for the service.
Make supplier information easier to review
A second scenario could organise delivery records and flag missing information for purchasing staff. The task might be to prepare a comparison of ordered and received items, using a defined set of documents. The system should identify uncertain matches rather than forcing every record into a neat total.
Keep the commercial decision with the authorised person. Identifying a discrepancy does not approve a new order or establish that a supplier is responsible. The output should give the reviewer enough context to investigate. A useful test includes partial deliveries and documents that use different item descriptions.
An operational pilot should support the person responsible for the service.
Use approved information for guest responses
A guest enquiry workflow needs current information from the venue. Opening times, booking conditions and event details should come from approved sources with an owner. When a request falls outside those sources, the workflow should direct it to the appropriate staff member.
Review questions involving allergies or special requirements with particular care. A language model should not infer ingredient safety from a promotional menu description. The operational design should route such a question to a responsible person with access to verified information.
Set boundaries for staff and guest data
A service preparation task should use only the personal information it needs. A staff member reviewing a delivery record may have no reason to access guest contact details. Define the roles and the data boundary before connecting the underlying systems.
The same discipline applies to external actions. A draft guest message, a proposed rota change and a supplier instruction each need their own authority rule. The team should know who can approve the action and how to stop or correct it. The operating venue remains responsible for its decisions.
Measure what the team can observe
For a service briefing pilot, measure the time required to prepare and review the result, along with the corrections needed before use. Check whether the draft retains important changes and identifies missing inputs. Avoid presenting preparation speed as evidence of better service without measuring that outcome.
Choose one team, one task and a manageable set of approved records. Agree what would justify continuing, revising or ending the pilot. That produces evidence the operator can use, while keeping the work close to the judgement and service responsibilities that make hospitality distinctive.
Explore the hospitality section of the My NEO Group network and discuss a specific operational pilot with the Group’s technology team.
