AI-powered hotel revenue management platform with real-time competitor tracking, PMS integration, and intelligent rate optimization.

Screenshots • Features • Tech Stack • Architecture • Chrome Extension • AI Engine • Security

AI-powered hotel revenue management — know exactly when to raise your rates. Real-time competitor scraping, PMS integration, and demand-aware pricing in a single platform.

Interactive 365-day rate canvas with color-coded AI recommendations. Bubble size reflects occupancy, colors indicate pricing action (green = raise, red = lower, purple = optimal). Real-time competitor data drives every suggestion.

Per-date rate management with AI-generated optimal pricing. View modes for day, week, month, and year with confidence scores and demand multipliers.

Revenue analytics with RevPAR, ADR, occupancy trends, health scores, and YoY comparisons. Smart alerts flag anomalies in booking velocity and competitor pricing.

Auto-discover competitors within 5km via Google Places. Real-time rate scraping from Google Hotels, Booking.com, and Expedia. Track pricing patterns, reviews, and market positioning.

Booking velocity snapshots with 14-day lookahead, pickup trends, and pace comparisons to identify demand surges early.

Historical pricing and occupancy trends with seasonality detection. Year-over-year pattern analysis for strategic rate planning.

Google review sentiment analysis, topic extraction, rating trends, response management, and AI-drafted replies — all from one dashboard.

Aggregate demand signals from 5 external sources (Ticketmaster, PredictHQ, Skiddle, Football Data, UK Bank Holidays) plus custom events. Demand multipliers automatically boost rate recommendations.

12-module operations suite: daily room status, staff management, reservation tracking, breakfast forecasting, linen management, finance/P&L, scheduling, and variance reporting.

Google Business Profile integration with business insights (views, calls, clicks), review analytics, and direct review reply from the dashboard.

Multi-provider auth (Email, Google, Facebook) with a 3-step hotel setup wizard including automatic competitor discovery.

• AI Rate Recommendations — per-date pricing using occupancy, competitor rates, events, weather, seasonality, and market positioning
• Interactive Bubble Canvas — 365-day calendar with physics-based visualization: bubble size = occupancy, color = pricing action
• View Modes — Today, Week, Month, Year with date navigation
• Rate Actions — Apply, A/B test, bulk approve, and historical tracking
• Confidence Scoring — recommendations only shown when 10+ competitor rates are available
• Demand Multiplier Engine — events + weather + seasonality compound into dynamic rate boosts (0.90x–1.35x)

• Auto-Discovery — Google Places API search within 5km radius
• Real-Time Scraping — Google Hotels, Booking.com, Expedia via SerpAPI with parallel browser pools (Playwright)
• Rate Monitoring — track competitor price changes over time with historical comparisons
• Review Integration — Google reviews with sentiment analysis, cached 24 hours
• Market Positioning — scatter-plot analysis of price vs. rating across the competitive set

• Multi-Turn Conversations — threaded chat with full hotel context
• Streaming Responses — Server-Sent Events (SSE) for real-time token rendering
• Tool Calling — AI autonomously executes: rate changes, event creation, web searches
• Persistent Memory — 8 categories (hotel profile, pricing strategy, competitor intel, guest feedback, market conditions, performance metrics, custom notes, learning history)
• Memory Decay — confidence scores decrease over time, requiring revalidation
• Web Search — Jina AI integration for real-time market intelligence
• Fallback Suggestions — contextual recommendations when API is unavailable

• Custom Events — create with type, date range, and impact level
• 5 External Sources — Ticketmaster, PredictHQ, Skiddle, Football Data API, UK Bank Holidays
• Demand Impact — automatic rate multipliers based on event proximity and magnitude
• Event Calendar — unified view of all demand signals with date filtering

• OAuth 2.0 — secure Google Business Profile connection
• Review Analytics — sentiment analysis, topic extraction, rating trends, response rates
• Business Insights — views, calls, direction requests, booking attribution, search keywords
• Review Replies — respond to reviews directly from the dashboard
• Auto-Refresh — background schedulers (24h for insights, 30min for reviews)

• Auto-Detection — activates on hopsoftware.com domains
• Data Scraping — rate calendar, occupancy percentages, reservation details with multi-page pagination
• Rate Apply — auto-fills AI-recommended rates into PMS forms
• Knockout.js Extraction — bypasses DOM to access PMS internal ViewModels directly
• MutationObserver — auto-syncs when DOM changes without user interaction
• Manifest V3 — modern security-focused architecture with Chrome storage

• Daily room status & staff assignments
• Staff directory, departments, scheduling, KPIs
• Reservation CRUD with PMS sync and bulk import
• Room inventory & off-service tracking
• Breakfast forecasting (covers, guests, food cost, revenue)
• Linen management with usage forecasting
• Finance module (budget, invoices, P&L, YTD)
• 7+ day scheduling with availability templates
• Variance tracking (actual vs. planned)
• Performance reports and operational metrics

• Revenue Analytics — monthly performance, RevPAR, ADR
• Occupancy Trends — historical patterns with YoY comparison
• Health Score — composite metric (occupancy + revenue + rating)
• Pickup Reports — booking velocity snapshots, 14-day lookahead
• Quick Stats — KPI cards with sparkline trend indicators
• Competitor Intelligence — pricing patterns, seasonality, positioning

• Authentication — Firebase email/password + Google OAuth + Facebook OAuth
• Hotel Setup Wizard — 3-step onboarding with automatic competitor discovery
• Dark/Light Theme — system-aware with manual toggle
• Admin Panel — user management, hotel suspension, platform announcements
• Rate Limiting — per-endpoint API protection via slowapi
• A/B Testing — built-in rate experiment framework with performance tracking

┌─────────────────────────────────────────────────────────────────┐
│                    CHROME EXTENSION (MV3)                        │
│  ┌─────────────┐  ┌──────────────┐  ┌───────────────────────┐  │
│  │ Content      │  │  Service     │  │  Popup UI             │  │
│  │ Scripts      │  │  Worker      │  │  Credentials + Sync   │  │
│  │ Hop Scraper  │  │  Background  │  │  Status + Controls    │  │
│  │ Rate Apply   │  │  Data Sync   │  │                       │  │
│  │ Overlay UI   │  │  Rate Fetch  │  │                       │  │
│  └──────┬──────┘  └──────┬───────┘  └───────────────────────┘  │
│         └────────────────┘                                      │
└──────────────────┬──────────────────────────────────────────────┘
                   │ HTTP + Chrome Storage
┌──────────────────┼──────────────────────────────────────────────┐
│             REACT FRONTEND (Vite + React Router)                │
│  ┌────────────┐  ┌────────────┐  ┌───────────┐  ┌──────────┐  │
│  │   Rate     │  │    AI      │  │ Analytics │  │ Housekpg │  │
│  │  Bubbles   │  │   Chat     │  │ & Reports │  │   12     │  │
│  │  Canvas    │  │   SSE      │  │  Charts   │  │ Modules  │  │
│  │  Physics   │  │  Stream    │  │  RevPAR   │  │          │  │
│  └─────┬──────┘  └─────┬──────┘  └─────┬─────┘  └────┬─────┘  │
│        └───────────────┼───────────────┼──────────────┘        │
│                        │  REST + SSE                            │
└────────────────────────┼────────────────────────────────────────┘
                         │
┌────────────────────────┼────────────────────────────────────────┐
│               FASTAPI BACKEND (234 Endpoints)                   │
│                        │                                        │
│  ┌─────────────────────┴─────────────────────┐                 │
│  │            9 API Routers                   │                 │
│  │  /auth  /hotels  /rates  /competitors      │                 │
│  │  /analytics  /scraper  /chat  /events      │                 │
│  │  /housekeeping  /admin  /pms  /occupancy   │                 │
│  └──┬───────┬────────┬────────┬──────┬───────┘                 │
│     │       │        │        │      │                          │
│  ┌──┴──┐ ┌──┴───┐ ┌──┴──┐ ┌──┴──┐ ┌─┴────┐                   │
│  │Rate │ │Scrap-│ │ AI  │ │Event│ │Google│                     │
│  │Gen. │ │ er   │ │Asst.│ │Mgr. │ │Biz.  │                    │
│  │     │ │Pool  │ │Tools│ │5 API│ │OAuth │                     │
│  │Demand│ │Play- │ │Mem. │ │Merge│ │Sched.│                    │
│  │Model │ │wright│ │Decay│ │     │ │      │                    │
│  └──┬──┘ └──┬───┘ └──┬──┘ └──┬──┘ └──┬───┘                    │
│     └───────┴────────┴───────┴───────┘                          │
│                      │                                          │
└──────────────────────┼──────────────────────────────────────────┘
                       │
        ┌──────────────┼──────────────┐
        │              │              │
   ┌────┴────┐   ┌────┴────┐   ┌────┴────────┐
   │ Firebase │   │   AI    │   │  External   │
   │Firestore │   │ Engine  │   │   APIs      │
   │  Auth    │   │  Groq   │   │ Google Maps │
   │ 10+ Coll │   │  Jina   │   │ SerpAPI     │
   │          │   │  Tools  │   │ Ticketmaster│
   │          │   │  Memory │   │ PredictHQ   │
   │          │   │         │   │ Open-Meteo  │
   └─────────┘   └─────────┘   └─────────────┘

                    ┌──────────────┐
                    │  PMS Sync    │  (Chrome Extension scrapes occupancy)
                    └──────┬───────┘
                           │
     ┌─────────────────────┼─────────────────────┐
     ▼                     ▼                     ▼
┌─────────┐         ┌──────────┐          ┌──────────┐
│Occupancy│         │Competitor│          │  Events  │
│  Data   │         │  Rates   │          │ + Weather│
│ per day │         │ Scraped  │          │ + Season │
└────┬────┘         └────┬─────┘          └────┬─────┘
     │                   │                     │
     └───────────────────┼─────────────────────┘
                         │
                    ┌────┴─────┐
                    │   Rate   │
                    │Generator │
                    │ AI Model │
                    └────┬─────┘
                         │
              ┌──────────┼──────────┐
              ▼          ▼          ▼
        ┌──────────┐ ┌───────┐ ┌────────┐
        │ Optimal  │ │  A/B  │ │ Bubble │
        │  Rate    │ │ Tests │ │ Canvas │
        │ per Date │ │       │ │ Render │
        └──────────┘ └───────┘ └────────┘

Service Layer Separation — 45+ backend services isolate business logic from API routing. Rate generation, scraping, AI chat, and event management each have dedicated service modules.

Parallel Browser Pools — Playwright manages 5 concurrent headless browsers for rate scraping, with orchestration for deduplication and error recovery.

Tool-Calling AI Agent — The AI assistant can autonomously execute backend actions (modify rates, create events, search the web) through a structured tool-calling interface with Groq LLM.

Memory Decay System — AI memory items have confidence scores that naturally decay over time, requiring periodic revalidation to prevent stale context.

Demand Compounding — Multiple demand signals (events, weather, seasonality, competitor pricing) are mathematically compounded into a single multiplier (0.90x–1.35x) that adjusts rate recommendations.

Firebase Transactions — Hotel creation, user linking, and rate updates use atomic Firestore transactions to prevent race conditions in multi-user environments.

Background Schedulers — asyncio-based tasks run autonomously: Google reviews (30min), business insights (24h), event cleanup (3AM), memory decay (4AM), PMS sync (15min).

The RatePulse Chrome Extension bridges the gap between the AI platform and hotel Property Management Systems.

┌─────────────────────────┐     ┌──────────────────────┐
│    Hop PMS Website      │     │   RatePulse Backend   │
│  pms.hopsoftware.com    │     │   /api/pms/sync       │
│                         │     │                        │
│  ┌───────────────────┐  │     │  ┌──────────────────┐ │
│  │  Rate Calendar    │──┼──▶──┼──│  Store Rates     │ │
│  │  Occupancy Grid   │  │     │  │  Generate AI     │ │
│  │  Reservations     │  │     │  │  Recommendations │ │
│  │  Knockout.js VM   │  │     │  │  Return Optimal  │ │
│  └───────────────────┘  │     │  └────────┬─────────┘ │
│                         │     │           │            │
│  ┌───────────────────┐  │     │           │            │
│  │  Auto-Fill Rates  │◀─┼──◀──┼───────────┘            │
│  │  (Apply Button)   │  │     │                        │
│  └───────────────────┘  │     │                        │
└─────────────────────────┘     └──────────────────────┘

Key Capabilities:

• Auto-activates on PMS domains (no manual trigger)
• Extracts data via DOM + Knockout.js ViewModel bypass
• Multi-page pagination with MutationObserver auto-detection
• Applies recommended rates back into PMS with one click
• Secure credential storage via Chrome Storage API
• 2-minute timeout for large data syncs
• Real-time sync status feedback

RatePulse's AI operates at three levels:

The core pricing algorithm combines multiple data sources into per-date recommendations:

• LLM: Groq API with multi-turn context
• Tool Calling: Rate modifications, event creation, web search execution
• Memory: 8-category persistent memory with confidence decay
• Streaming: SSE for real-time response rendering
• Context: Full hotel profile, recent rates, competitors, events injected per message

• Velocity-based booking detection — flags unusual pickup patterns (critical/high/medium/low)
• Demand spike alerts — threshold-based notifications for occupancy surges
• Competitor price changes — monitors and alerts on significant rate movements

ratepulse/
├── src/                          # React Frontend
│   ├── App.jsx                   # Router with auth guards
│   ├── HotelRateBubbles.jsx      # Main dashboard (4,074 lines)
│   ├── components/               # 46 React components
│   │   ├── AIChat.jsx            # Revenue assistant (3,400 lines)
│   │   ├── tabs/                 # 8 main feature tabs
│   │   ├── auth/                 # Login, Register, AuthProvider
│   │   ├── admin/                # Admin panel
│   │   └── modals/               # Rate actions, competitors, settings
│   └── services/                 # API client layer
├── backend/
│   ├── main.py                   # FastAPI app entry
│   ├── config.py                 # Environment configuration
│   ├── routers/                  # 9 API routers (234 endpoints)
│   │   ├── auth.py               # Authentication
│   │   ├── hotels.py             # Hotel CRUD & settings
│   │   ├── rates.py              # Rate generation & actions
│   │   ├── competitors.py        # Competitor tracking
│   │   ├── analytics.py          # Revenue analytics
│   │   ├── scraper.py            # Rate scraping orchestration
│   │   ├── chat.py               # AI assistant (85 endpoints)
│   │   ├── events.py             # Event management
│   │   └── housekeeping.py       # Operations
│   ├── services/                 # 45+ business logic modules
│   │   ├── rate_generator.py     # AI pricing engine (43KB)
│   │   ├── chat.py               # Conversation management (36KB)
│   │   ├── external_events.py    # 5-source event aggregation (49KB)
│   │   ├── smart_alerts.py       # Anomaly detection (33KB)
│   │   ├── housekeeping.py       # 12-module ops engine (61KB)
│   │   ├── scraper/              # Playwright browser pools
│   │   └── schedulers/           # Background tasks
│   └── models/                   # 10 Pydantic model files
├── extension/
│   ├── manifest.json             # Chrome MV3 config
│   ├── background.js             # Service worker
│   ├── content/
│   │   ├── hop-scraper.js        # PMS data extraction (62KB)
│   │   └── overlay.js            # UI overlay (33KB)
│   └── popup/                    # Extension popup UI
├── vite.config.js                # Build configuration
└── package.json                  # Frontend dependencies

• Parallel Scraping — 5 concurrent Playwright browsers for competitor rate extraction
• In-Memory Caching — TTL caches for API responses, competitor data, and rate calculations
• SSE Streaming — AI responses render token-by-token, no waiting for completion
• Background Schedulers — reviews (30min), insights (24h), cleanup (3AM), decay (4AM), PMS sync (15min)
• Vite Production Build — tree-shaking, code splitting, and optimized asset bundling
• Firebase Batch Operations — bulk writes for rate updates and occupancy imports
• MutationObserver — extension auto-syncs without polling, zero CPU waste

Built by xelauvas.dev

_{This repository contains only documentation. Source code is proprietary.}