Skip to content
HotelMind AI

Business Analysis

User Personas

Six personas mapped to real and proposed roles - representative, not customer-validated.

Source: docs/product-management/04-user-personas.md

HotelMind AI — User Personas

These are representative/proposed personas constructed from the product's implemented role model and general hospitality-industry job-function knowledge — not derived from customer interviews, surveys, or usage analytics, since no such data collection exists yet for HotelMind AI. Each persona notes which backend role, if any, it maps to.


General Manager — "Maria"

Proposed persona; not a distinct backend role (accessible via the Owner role or the display-only Perspective selector). Oversees all property departments and is accountable to ownership for overall performance. Job to be done: "When I start my day, I want one view of occupancy, revenue, and any flagged issues, so I can decide where to focus." Needs the Executive Dashboard and AI Insights, with clear labeling of which figures are real versus simulated — already implemented via the DataSourceBadge, since Maria will act on this data.

Revenue Manager — "David" (REVENUE_MANAGER)

Owns pricing strategy for room inventory. Job to be done: "When I set tomorrow's room rate, I want a demand forecast and a pricing recommendation with rationale, so I can decide quickly and defensibly." David's real frustration: occupancy forecast accuracy is currently weak (MAPE 72–120% per internal validation), so he needs to know when to trust versus discount a recommendation. Confidence/uncertainty must be visible, not hidden behind a single point forecast.

Operations Manager — "Priya" (OPS_MANAGER)

Coordinates across rooms, restaurant, and staff to keep daily operations running. Job to be done: "When I plan tomorrow's staffing, I want a demand-informed recommendation, so I don't over- or under-staff." Priya is aware, via internal documentation, that current restaurant and staffing recommendations are trained on synthetic — not real — operational data, and needs that caveat surfaced, not hidden. Success means recommendations become more trustworthy as real data replaces synthetic data over time.

Restaurant Manager — "Carlos" (RESTAURANT_MANAGER)

Manages the hotel restaurant's daily service and food cost. Job to be done: "When I plan tomorrow's prep, I want a breakfast/lunch/dinner demand forecast, so I reduce waste without running out." No real restaurant order data currently exists in the system — forecasts are trained entirely on synthetic data, a significant caveat Carlos needs to weigh, made visible through the beta data-source badge.

HR / Staff Manager — "Angela" (proposed persona)

Not an implemented backend role — represented today only through the mock-data-backed HR & Payroll page. Job to be done: "When I build next week's schedule, I want a staffing-level recommendation, so I can staff efficiently." Angela's most-needed feature — automated shift scheduling — is explicitly unimplemented (the OR-Tools scheduler raises NotImplementedError). She should not be presented with the HR page as if it were a working feature.

Front Desk Manager — "Tom" (proposed persona)

Not an implemented backend role — the Check-in/Check-out page is mock-data-backed. Job to be done: "When a guest arrives, I want to check them in against their reservation quickly." This workflow happens outside HotelMind AI today (or via a separate PMS); the in-app page is a design preview, not a production workflow, until backend integration is complete.