The Brain
Atlas · Bay Street Hospitality
LIVE · 14:38 SGT 6 source layers · 10 functions · 14 assets streaming

The decision-making engine between operations and capital.

Atlas is not software. It is the control layer that converts hotel data into intelligence, intelligence into decisions, and decisions into returns — closing the loop between operating teams, market signals, and investment strategy.

Source layers connected
6
▲ all online
Signals processed · 24h
847,290
▲ +18% vs avg
Decisions surfaced · 24h
142
▲ 23 high-priority
Decisions executed · 24h
94
▲ 66% execution rate
NOI lift attributed · LTM
$6.4M
▲ across 14 assets
Connection Layer · Inputs · what feeds the brain
LIVE
PMS · RMS · POS
Reservations, occupancy, ADR, RevPAR, F&B
Opera Cloud + IDeaS 14 assets
LIVE
Financial · ERP
Cash flow, capex, debt schedules, GL
NetSuite + Yardi Daily sync
LIVE
External Market
STR comp set, flight capacity, FX, macro
STR + Bloomberg 170 countries
LIVE
Investment Engine
Bay Score, BMRI, IRR forecasts, Z-scores
Atlas internal 15 chains
LIVE
Capital · Financing
Lender terms, debt benchmarks, CMBS rates
9 lender APIs Quarterly
LIVE
Operator Layer
Asset managers, GMs, revenue managers
14 GMs · 9 AMs Real-time

Atlas Brain

10 transformation functions · operations × capital
1.2M
compute · last 24h
FN 1
Normalize & Unify
Single operating + financial truth across properties
FN 2
Translate Ops → Returns
Occupancy + ADR → real-time IRR & yield
FN 3
Forward-Looking Forecasting
Macro + demand + comp set → revenue projections
CORE MOAT
FN 4
Decision Engine
Pricing, capex, repositioning · prescriptive output
MOAT
FN 5
Execution Layer
Pushes decisions into RMS, PMS, operator tasks
FN 6
Portfolio Optimization
Cross-asset capital + attention allocation
FN 7
Capital Allocation
Buy / hold / sell / refi / develop signals
FN 8
Credit Intelligence
Optimal leverage, debt pricing, risk-adj. credit
MOAT
FN 9
Continuous Learning
Forecast vs actual → model refinement
FN 10
Institutional Reporting
IC memos, LP dashboards, performance reports
Execution Layer · Outputs · where decisions land
PRICING
RMS Pricing Push
+$28 ADR
Cedar Tokyo · executed 14:34 SGT via Opera Cloud
FN 4 → IDeaS RMS
CAPITAL
Refi Trigger Signal
−85 bps
Aurora · refi window optimal · 30-day execution lead
FN 7 + FN 8 → Treasury
OPERATIONS
Capex Timing Signal
Defer 90d
Glacier HIE Manchester · MEES retrofit window
FN 4 → AM workflow
REPORTING
Auto-IC Memo
36 pages
Project Sandstone · ready for IC review
FN 10 → LP portal
Live Decision Feed STREAMING
PRICING · EXECUTED 14:34 SGT
Cedar Tokyo · ADR +$28 recommended · auto-pushed to RMS
PMS+STR → FN 4 → IDeaS RMS
NOI Δ proj. (annualized) +$420K
FLAG · SURFACED 14:31 SGT
Aurora · margin compression flag · operator review queued
ERP+POS → FN 9 → AM dashboard
OpEx variance +250 bps vs UW
CAPITAL · QUEUED 14:28 SGT
Aurora · refi window opens · CMBS pricing favorable −85 bps
9 lender feeds → FN 7+8 → Treasury
Annual cost saving $13.6M
PRICING · PENDING 14:22 SGT
Marunouchi Capital · channel-mix shift · OTA share 38% → 24%
PMS+CRS → FN 4 → SiteMinder
RevPAR lift proj. +90 bps
OPS · DISPATCHED 14:15 SGT
Glacier HIE Manchester · MEES retrofit · CapEx defer 90 days
Macro+Capex → FN 6 → AM workflow
Cash preservation $2.1M Y1
PRICING · EXECUTED 14:08 SGT
Mayfair House · weekday ADR optimization · +£42 effective
PMS+STR → FN 4 → IDeaS RMS
NOI Δ proj. +£312K
FLAG · SURFACED 13:54 SGT
Bondi Crescent · forecast vs actual gap +280 bps · investigate
PMS+ERP → FN 9 → AM dashboard
Variance type Persistent (3mo)
CAPITAL · SIGNALED 13:42 SGT
Tide · partial-disposition modeled · 40% stake H2 2026
All layers → FN 7 → IC queue
Capital recycling $480M
OPS · LEARNING 13:28 SGT
Tide playbook · validated +340 bps NOI lift · template extracted
Forecast vs actual → FN 9 → Model
Apply to 3 candidates

Three conversion loops.

Atlas is structured around three sequential transformations. Each compounds: better data raises the floor on intelligence; better intelligence sharpens decisions; sharper decisions improve the actuals fed back into the system.

LOOP 01

Data → Intelligence

Raw data→Insight
Hotel + macro inputs become normalized, contextualized insight. The output is no longer "occupancy was 78%" — it's "occupancy was 78%, vs comp set at 81%, vs forecast at 76%, with implied RevPAR yield gap of 220 bps."
LOOP 02

Intelligence → Decisions

Insight→Action
Insight becomes prescription: pricing moves, capex timing, capital reallocation, financing structure. The brain doesn't report state — it issues instructions with magnitudes, confidence intervals, and execution paths.
LOOP 03

Decisions → Returns

Action→IRR
Executed decisions show up in actuals: NOI lift, IRR realization, cost-of-capital improvement. The feedback closes via FN 9 — every decision's outcome trains the model that produced it, compounding the system's edge over time.

Defensibility · the moat.

The brain's edge is not any single capability. It is the combination — and the fact that legacy systems can't replicate the closed loop without rebuilding from the data layer up.

What legacy systems do

Reporting & dashboards

  • Static reporting — yesterday's data, presented in panels.
  • Single-property view — limited cross-asset learning.
  • Operator-only context — no capital-side integration.
  • Disconnected from execution — insights require human translation into action.
  • No feedback loop — forecasts never improve from outcomes.
What Atlas does

Closed-loop decisioning

  • Cross-market learning — every asset's outcomes train every other asset's models.
  • Macro-integrated forecasting — sovereign signals + tourism flows + comp set in one model.
  • Capital-aware — pricing decisions know cost of capital; financing decisions know operating reality.
  • Closed-loop execution — decisions push to RMS, PMS, operator workflows directly.
  • Compounding intelligence — every decision's actual outcome refines the model that produced it.
Bottom line. Atlas is the decision-making engine that sits between operations and capital. It connects hotel systems, market data, financials, and investors — and turns all of it into real-time, executable strategies that maximize returns. Most systems do reporting. Atlas does control.