Tool 01
Day by Day Strategy
Case Study · Hotel Switchboard LLC · RevEVOLVE · Feb 2026
The First
AI-Native
Revenue Manager
RM Copilot eliminates 60–70% of revenue management's administrative grind — giving every hotel RM the intelligence of a 15-year veteran, available 24/7 across every property in your portfolio.
0%
of RM time lost to
repetitive data tasks
repetitive data tasks
+0%
RevPAR lift at
full deployment
full deployment
0+
Properties per RM
(up from 12)
(up from 12)
$0K
Net annual savings
per portfolio
per portfolio
Section 01
The Strategic
Bandwidth Crisis
Revenue Managers are the highest-leverage role in a hotel — yet most spend the majority of their time on repetitive data work that should be automated, not strategized.
Where RM Time Actually Goes
The 20–30% that drives 80% of revenue gets squeezed because the other 70% is unavoidable — until now.
Missed pricing windowsDemand spikes go unnoticed until it's too late. Competitors capture the booking before data is even pulled.
Reactive, not proactiveWithout continuous monitoring, RMs respond to market events after they happen — never before them.
RM burnout at scaleManaging 10+ properties manually is unsustainable. Most RMs burn out within 18 months of exceeding that threshold.
Linear scaling bottleneckEvery new property means hiring another RM. Revenue management headcount grows proportionally — destroying gross margins.
Section 02
The Architecture
Behind RM Copilot
Built on Model Context Protocol (MCP), each data domain becomes a tool the Claude AI agent invokes dynamically — pulling only what's needed, when it's needed, for consistent and contextually accurate responses.
Claude AI
RM Persona + 9 Glossaries
MCP Server
RevEVOLVE Custom
9 Live Data Tools
PMS · RMS · STR · Events
5 Interaction Modes
Summary · Chat · Voice
Tool 02
Monthly Summary
Tool 03
Market Segments
Tool 04
Booking Window
Tool 05
Rate Movement
Tool 06
Pricing Ladder
Tool 07
Reservation Activity
Tool 08
Weekly STR
Tool 09
Events by Date
Section 03
12 Signal Rules
That Never Sleep
The Signal Engine scans all data streams continuously, firing prioritized alerts with dollar-impact quantification before the RM would ever notice the problem manually. Click any signal to see its deep dive.
CRITICALRate UnderpricingSelf < comp >10% & OCC >50%
CRITICALBehind PaceOTB < Forecast >30% & <14d
CRITICALOOO SpikeOOO rooms >20% inventory
HIGHDemand Spike7d pickup >2× 8wk avg DOW
HIGHADR DilutionOTA Discount >30% & gap >$30
HIGHGroup Wash RiskBlock >10 & pickup <50% <7d
HIGHWeekend Rate ErosionFri/Sat ADR < LY >15%
HIGHRate Parity ViolationSelf rate differs >5% channels
HIGHSegment LeakSegment ADR < avg >30% & >10%
MEDIUMEvent ProximityEvent within 7d & pace < STLY
MEDIUMBudget GapMonth OTB rev < Budget >20%
MEDIUMLead Time ShiftAvg lead time ↓ >3d vs STLY
CRITICAL
Rate Underpricing — Valentine's Friday
+$589
incremental revenue per night at $139 BAR
BAR $120 is the lowest in the competitive set. The hotel is at 71% OTB with +12 rooms pickup in the last 7 days. Comp set average: $139. This is a live Valentine's Day demand event — every hour at $120 is revenue left on the table.
Recommended Actions
Raise BAR to $139 immediately
Close OTA Discount below $119
Push to $149 if OTB reaches >80%
Section 04
Five Interaction Modes
RM Copilot meets every stakeholder wherever they work — all five modes powered by the same unified MCP data layer for consistent, contextually accurate answers every time.
Daily Summary
Auto-generated morning briefs with KPI snapshots, 7-day outlook, and a ranked signal panel — delivered to email and dashboard every day.
Every 6 AM · Morning Scan
Signal Deep Dive
Click any signal to dynamically pull supporting data, historical context, what-if scenario modeling, and specific recommended actions.
On-Demand · Any Signal
Chat
Natural language Q&A and what-if analysis via web and mobile. Ask anything — RM Copilot queries live data to respond with precision.
Web + Mobile · Ad Hoc
Voice Call
Phone or WebRTC voice conversation with sub-2-second end-to-end latency. Hands-free, full context, post-call summary auto-generated.
Call Anytime · <2s Latency
Scheduled Meeting
AI-facilitated multi-stakeholder revenue calls — auto-agenda, real-time Q&A, live action capture, and full post-meeting pipeline.
AI-Facilitated · Weekly
Section 05
The Business Case
for RM Copilot
A real-world scenario: a hotel management company with 28 properties and 3 Revenue Managers — before and after deploying RM Copilot.
Without RM Copilot
28 properties
Total portfolio size
3 Revenue Managers
Headcount
$285K / RM
Revenue per RM
Must hire RM #4
8 new properties added
$95K/year
RM #4 salary + benefits
73%
Gross margin
With RM Copilot
36 properties
Same 3 RMs — expanded portfolio
3 Revenue Managers
Headcount unchanged
$425K / RM
Revenue per RM (+49%)
RM #4 avoided
Same team, 8 more properties
$40K/year
Technology investment
81%
Gross margin (+8 pts)
Phase 1 → Phase 4 Success Metrics
Section 06
The Competitive Moat
Legacy RM platforms — IDeaS, Duetto, Lighthouse RMS, Atomize — were built as pricing engines in a pre-LLM world. RM Copilot attacks the $3.2B market from an angle they are architecturally unable to replicate.
| Capability | IDeaS / Duetto | Lighthouse RMS | Atomize | Generic AI | RM Copilot |
|---|---|---|---|---|---|
| Deep RM Domain Knowledge | Partial | Partial | ✕ | ✕ | ✓✓ 199 terms + 12 rules |
| Proactive Signal Detection | Basic alerts | Basic alerts | Basic alerts | ✕ | ✓✓ 12-rule engine |
| Rate Parity & Comp Monitoring | ✓ | ✓ Core strength | ✓ | ✕ | ✓✓ Via Rate Movement tool |
| Natural Language Chat | ✕ | ✕ | ✕ | Generic only | ✓✓ RM-specialized |
| Voice Call Interface | ✕ | ✕ | ✕ | ✕ | ✓✓ WebRTC + PSTN <2s |
| AI Meeting Facilitator | ✕ | ✕ | ✕ | ✕ | ✓✓ Full pipeline |
| What-If Revenue Modeling | Partial | Partial | ✕ | Limited | ✓✓ Pricing ladder model |
| Daily AI Brief + Email | ✕ | Reports only | ✕ | ✕ | ✓✓ 6 AM AI-generated |
| Unified 9-Source Data Layer | Partial | Rate + STR data | Partial | ✕ | ✓✓ MCP unified layer |
Deep Domain Expertise
199 hotel RM terms, formulas, and synonyms — plus 12 signal rules encoding decades of practitioner wisdom — loaded directly into the agent's system prompt. Generic AI cannot replicate this without years of hospitality-specific development.
Data Integration Moat
The RevEVOLVE MCP server unifies 9 data sources — PMS, RMS, STR, rate shopping, events — into one consistent, real-time layer. A competitor starting today needs 12–18 months to build comparable integrations.
Multi-Modal Interaction
No competitor offers Daily Summary + Chat + Voice + AI Meeting Facilitation in a single RM agent. IDeaS, Duetto, Lighthouse, and Atomize are pricing engines — tools you query. RM Copilot is a revenue management colleague that proactively tells you what to do.
Section 07
12-Month Implementation Roadmap
A phased build that generates pilot revenue from Month 1 and compounds capability through Month 12.
Phase 1
Foundation
Months 1–3
- MCP Server (9 endpoints)
- Signal Engine (12 rules)
- Daily Summary dashboard
- Email delivery pipeline
- PDF export
- 3–5 pilot properties
Phase 2
Conversational AI
Months 3–6
- Chat (web + mobile)
- Signal Deep Dive
- What-if analysis
- Conversation memory
- Cross-session context
- 15–20 properties
Phase 3
Voice Integration
Months 6–9
- WebRTC in-app voice
- PSTN via Twilio
- STT → LLM → TTS (<2s)
- Post-call auto-summary
- Beta: 5–10 GMs
- Voice action capture
Phase 4
Meeting Facilitator
Months 9–12
- Multi-party WebRTC
- Auto-agenda engine
- Participant ID + actions
- Visual companion
- Post-meeting pipeline
- Commercial launch
Section 08
Frequently Asked Questions
Everything revenue managers, hotel ownership groups, and management companies need to know about RM Copilot and AI-powered hotel revenue management.
RM Copilot is an AI-powered Revenue Management Agent built by Hotel Switchboard LLC under the RevEVOLVE platform. It is designed for hotel Revenue Managers and hotel management companies managing multi-property portfolios. Rather than replacing RMs, RM Copilot acts as a senior revenue management colleague — handling the 60–70% of repetitive data tasks so that human RMs can focus on the 20–30% of strategic decisions that drive 80% of revenue.
IDeaS, Duetto, Lighthouse RMS, and Atomize are pricing engines — powerful tools for rate optimization but built on a query-and-dashboard model. You ask them questions; they show data. RM Copilot operates differently: it is a proactive AI agent powered by Model Context Protocol (MCP) that continuously monitors 9 data streams, fires 12 signal detection rules when revenue-critical events occur, and communicates findings through 5 interaction modes including natural language chat, voice calls, and AI-facilitated revenue meetings. No competitor currently offers this combination.
Based on the modeled business case for a 28-property portfolio with 3 Revenue Managers: deploying RM Copilot at $40,000/year avoids a $95,000 RM hire while expanding the portfolio to 36 properties. This delivers a net saving of $55,000/year, raises revenue per RM from $285K to $425K (+49%), and improves gross margin from 73% to 81% (+8 percentage points). At Phase 4 deployment, properties can expect a +5–8% RevPAR lift versus control properties.
RM Copilot connects to 9 data domains via the RevEVOLVE MCP Server: (1) Day by Day Strategy — OCC, ADR, RevPAR, pickup, forecast, comp rates; (2) Monthly Summary — OTB, STLY, budget, forecast; (3) Market Segments — CY/LY room nights, revenue, ADR by segment; (4) Booking Window — RMS, ADR, revenue by lead time; (5) Rate Movement — self rate and competitor rates by date; (6) Pricing Ladder — room nights and avg rate by OCC slab; (7) Reservation Activity — rate, segment, channel per reservation; (8) Weekly STR — OCC/ADR/RevPAR self, comp set, index; (9) Events by Date — event name, type, dates, category tags.
RM Copilot is powered by Claude AI (Anthropic) as its reasoning and response engine, operating via the Model Context Protocol (MCP) architecture. The system prompt loads 9 domain-specific glossaries totalling 199 hotel revenue management terms with formulas and synonyms, plus an RM Persona that gives the agent the communication style and strategic priorities of a 15-year veteran Revenue Manager. The unified MCP data layer ensures consistent, contextually accurate answers across all 5 interaction modes.
The Signal Engine runs 12 detection rules continuously across all incoming data streams. Rules are classified as Critical (e.g., Rate Underpricing when self rate is more than 10% below comp avg and OCC exceeds 50%), High (e.g., Demand Spike when 7-day pickup exceeds 2× the 8-week average for the same day of week), or Medium (e.g., Budget Gap when monthly OTB revenue is more than 20% below budget). Each fired signal includes a plain-language description, dollar-impact quantification, supporting data, historical context, and specific recommended actions with owner and deadline.
RM Copilot follows a 4-phase, 12-month implementation roadmap. Phase 1 (Months 1–3) delivers the Foundation: MCP Server, Signal Engine, Daily Summary dashboard, and email delivery with 3–5 pilot properties live. Phase 2 (Months 3–6) adds Conversational AI: chat, Signal Deep Dive, and what-if analysis for 15–20 properties. Phase 3 (Months 6–9) integrates Voice via WebRTC and Twilio. Phase 4 (Months 9–12) launches the AI Meeting Facilitator for full commercial deployment. Pilot properties receive value from Day 1 of Phase 1.
RevEVOLVE (revevolve.ai) is a hospitality AI platform built by Hotel Switchboard LLC, co-founded and led by Harry Sheta (CEO). The company specializes in AI-native revenue management solutions for the hotel industry, with a focus on multi-property management companies. RM Copilot is RevEVOLVE's flagship product — an AI agent that functions as a senior Revenue Manager across every hotel in a portfolio.
Section 09 · Conclusion
The Question Is No Longer
Whether AI Transforms RM
The hotel industry has spent two decades building better dashboards and smarter pricing engines. RM Copilot makes a fundamentally different bet: the greatest productivity gain doesn't come from better tools, but from a trusted AI colleague that handles the work so human RMs can focus entirely on strategy.
"For hotel management companies, the question is not whether AI will transform revenue management. The question is whether they adopt RM Copilot before their competitors do."