Physics-aware AI for cooling optimization of buildings.

An optimization layer over your existing BMS. Continuously identifies lower-energy operating strategies, targeting up to 15% less cooling energy — no replacement of existing HVAC or BMS infrastructure, with comfort held as a hard constraint.

The platform

Intelligence for the plant you already run.

EnthalpyX connects to your BMS and continuously identifies a lower-energy way to meet the cooling demand you actually have — not the design-day the schedules were written for — adapting as weather and load change.

  • No major mechanical upgrades. Software-led — no replacement of your existing HVAC or BMS infrastructure.
  • Native BMS integration. BACnet/IP, Modbus and Tridium Niagara.
  • Edge + cloud architecture. A lightweight on-site gateway reads plant data and connects outbound to our cloud. No inbound ports, no direct exposure of the BMS.
  • Fail-safe by design. If the connection drops, your BMS reverts to its own schedules. Nothing in the building depends on us being online.
EnthalpyX AI optimization layer

Learns the building, identifies lower-energy operating strategies

↓ optimized setpoints  ·  ↑ live operating data
Already on site Your existing BMS

Remains in full control of execution, exactly as today

Already on site Your existing equipment

Chillers, pumps, cooling towers, AHUs — untouched

Under the hood

The predictive power of AI, held to the rigor of physics.

EnthalpyX models the cooling system as one connected thermal system and continuously identifies lower-energy operating strategies while respecting comfort, equipment, and operational limits.

System view
Your cooling plant as one connected system

EnthalpyX learns how your specific building responds — not how an average building responds — and adapts as conditions change.

Control variables
The levers your plant already has

Chilled-water supply temperature reset, condenser-water setpoints, chiller staging and sequencing, pump speeds, and pre-cool windows.

Optimization
A better way to meet the same cooling demand

Instead of holding to a fixed schedule, EnthalpyX continuously identifies a lower-energy operating strategy for the conditions you’re in right now.

Constraints
Physics-bounded, always

Every recommendation is checked against equipment operating envelopes and the comfort and safety limits your team defines. A strategy that violates either is never surfaced.

M&V
Savings measured against an agreed baseline

Pilot measurement and verification follows IPMVP principles, with the baseline and measurement boundary agreed with your team before optimization begins.

Working with us

Autonomy you adopt step by step.

Control is earned in three stages — and your team decides when to move to the next one.

01 ReadRead-only

We connect, build the baseline, and validate the model against your operating data. You see where the waste is and what it’s worth before anything changes.

02 AdvisoryOperator approved

The platform generates optimized setpoints; your operators review and execute them. Savings are measured against the agreed baseline as you go.

03 Closed loopBounded automation

Fully automated optimization inside operational boundaries your team defines — continuous, and never outside the lines you drew.

AI safety, by design
Human-defined boundaries
Physics bounded
BMS controlled
Fail safe
Concept Interface · Sample Data
app.enthalpyx.com / sites / building-a / overviewrecommendationsbaselinealertssettings
EnthalpyX
Building A Advisory Mode
Cooling Load · live
712kW
−14% vs. last-week typical
Estimated Savings · last 24h
$4,820
−13% vs. baseline
Comfort Range · occupied floors
Temp 22.4°C ✓
RH 51% ✓
attn Zone 14 sensor reading 0.6°C above neighbours. Flagged for review. 14:22 today
info Ambient +3°C forecast tomorrow. Pre-cool window suggested 13:30–14:30. 10:05 today
Open Recommendations · awaiting operator review

Each recommendation includes projected impact and a confidence level. In advisory mode, operators review and choose whether to send setpoints to the BMS.

Reset chilled-water supply from 6.5°C → 7.2°C
Cooling load and dew-point forecast support a higher CHWS temperature for the next 4 hours without breaching latent-load limits on floors 3–18.
Projected impact · −4.1% chiller kW Confidence · High Window · 14:30 – 18:30
Send to BMS Details
Stage down Chiller 2 during 02:00 – 05:00
Predicted overnight load falls well within Chiller 1 + Chiller 3 combined capacity. Removes one compressor start-stop cycle from Chiller 2.
Projected impact · −96 kWh Confidence · High Window · Tonight
Send to BMS Details
Raise condenser-water approach setpoint by 0.5°C
Cooling-tower fan power reduction offsets marginal chiller efficiency loss at current wet-bulb of 24.1°C. Tower fan speed drops to ~62% average.
Projected impact · −2.3% plant kW Confidence · Medium Window · Continuous
Send to BMS Details
Pre-cool floors 8–14 · 12:45 – 13:45
Ambient forecast +3°C tomorrow with peak tariff window 14:00 – 17:00. Shifting ~78 kWh out of peak reduces cost and flattens the afternoon load curve.
Projected impact · $310 tariff shift Confidence · Medium Window · Tomorrow
Send to BMS Details
IPMVP-aligned baseline · locked before optimization

Baseline agreed with the operations team before any setpoint changes. Independent variables track weather and occupancy so savings reflect optimization — not weather that would have driven the bill lower anyway.

Baseline period
01 Jan – 31 Mar 2026
90 days · 15-min data
Baseline locked
12 Apr 2026 · 09:14
Signed off by operations lead
Independent variables
OAT · RH · Occupancy
Weather from ADIA + on-site sensors
Model fit · CVRMSE
12.3%
Within ASHRAE Guideline 14 limit
Measurement boundary
Central plant only
Chillers, CHW pumps, CDW pumps, tower fans
Blended tariff assumption
$0.121 / kWh
DEWA peak-adjusted
Cumulative verified savings · last 30 days18,940 kWh · $2,290
Wk 1
3.9 MWh
Wk 2
5.3 MWh
Wk 3
4.5 MWh
Wk 4
5.2 MWh
Alerts · last 24 hours

Anything that touches comfort, equipment envelopes, sensor health or data quality is surfaced here — well before it shows up on a complaint log.

warn Chiller 2 evaporator approach trending 0.8°C above 30-day median. Tube fouling suspected — recommend inspection at next planned shutdown. 15:41 today
attn Zone 14 sensor reading 0.6°C above neighbours. Flagged for review. 14:22 today
attn CHW pump P-3 VSD holding at 100% for > 40 min. Investigate secondary loop bypass valve. 13:08 today
info Ambient +3°C forecast tomorrow. Pre-cool window suggested 13:30 – 14:30. 10:05 today
info Setpoint change queued for operator review · CHWS 6.5°C → 7.2°C. 09:47 today
ok Overnight staging recommendation executed as reviewed. −112 kWh vs. baseline. 06:02 today
ok Daily BMS heartbeat check · all points reporting nominally. 00:15 today
Operating boundaries · defined by your team

The platform never optimizes outside these limits. Changes are logged and require operator confirmation.

Occupied temperature range
Applied to floors 3 – 18 during occupied hours
21.5°C – 23.5°C
Locked
Relative humidity range
Latent-load constraint · GCC coastal profile
40% – 60%
Locked
CHW supply temperature bounds
Chiller efficiency envelope
5.5°C – 8.0°C
Locked
Minimum chiller run-time
Compressor short-cycle protection
45 minutes
Locked
Operating mode
Advisory · guard-railed · closed loop
Advisory (read-only)
Locked
BMS connection
BACnet/IP via edge gateway · outbound only
Gateway EG-041 · uptime 99.98%
Live
Notification recipients
Alerts routed to operations rota
3 recipients · ops-desk on-call
Locked

Illustrative UI with sample data — not a live deployment. Every real engagement starts in advisory mode.

What changes

Less energy in. Nothing taken away.

On the bill
  • Targeting up to 15% less cooling energyReduction target for plants running static schedules — measured against an agreed baseline per pilot
  • Software-led payback profileBecause there’s no mechanical retrofit, savings begin from deployment rather than after a capex cycle
  • Higher NOI potentialRecurring savings flow to operating income once verified in your building
In daily operations
  • Comfort is a constraint, not a trade-offComfort limits are enforced before any saving is counted
  • Gentler on the plantSmoother staging and fewer cycling events reduce mechanical stress
  • Nothing new to maintainNo boxes on the wall, no extra maintenance contract
In your reporting
  • Lower carbon and water footprintLess compressor work and less cooling-tower make-up water
  • Documentation designed for reviewBaseline, methodology and results structured to support ESG reporting and third-party review

Where we work

Engineered where cooling is the whole game.

In the GCC, cooling isn’t a season — it’s the operating condition. EnthalpyX is built for 45–50 °C ambient temperatures, high-humidity coastal profiles, and buildings where cooling dominates the energy bill year-round. Optimization tuned on mild climates doesn’t transfer here. Ours was never anywhere else.

45–50 °C Peak ambient conditions the models are engineered for
Year-round Cooling load profile — no shoulder-season shortcuts
High humidity Coastal latent loads handled explicitly, not averaged away

The first step

We start read-only.

Every engagement begins in advisory mode: we connect to your BMS, build a documented baseline, and show you where the savings are expected to be — before a single setpoint changes. If the evidence isn’t there, you’ll know early, and so will we.

EnthalpyX Physics-aware AI for cooling optimization of buildings.
© 2026 EnthalpyX. All rights reserved.