Cinematic upward shot of a tall glass-and-steel commercial building at night, with a constellation of warmly lit windows.
In validation · 2026

Catch the issue before the rebuild.

The miss

Periodic inspections + reactive tickets miss the slow stuff. By the time it's visible, it's expensive.

Slow leaks

Discovered after the rebuild bill.

Cracks that grow

Hairlines turn structural quietly.

Rebar corrosion

Cover + humidity + history compound.

Stress isn't uniform

One side bakes. Another stays damp.

Maintenance should be answered zone by zone — not whole-building.

How it works

01

Ingest

BMS · sensors · weather · specs · history.

no rip & replace
02

Engine

Rules + physics today. ML calibrates as outcomes accumulate.

explainable
03

Output

Score · ranked risks · next maintenance window.

where · what · when

Worked example

Cross-section of a building with the wet riser zone on floor 12 highlighted.
B-001 · F12-NORTHScore 62 ↓ from 81

Risk brief

Water ingress · F12-N

  • Where

    F12-NORTH · HW-RISER-12

  • Risk

    Water ingress + corrosion

  • Window

    Inspect in 7–14 days

  • Skip

    Whole-building waterproofing

Evidence

  • Indoor RH (F12-N)↑ 6 weeks
  • Rain events5 / 30d
  • Rebar cover25–35 mm
  • HistoryPrior leak, same riser
Try it on your buildings

The dashboard

Every morning, one screen. Where, what, when.

healthyinfra · portfolio B-00109:14 AM

Portfolio

Zones · 6 of 24

  • F11-S
    88
  • F12-N
    62
  • F14-E
    81
  • B1-W
    70
  • F08-N
    92
  • F03-S
    76

Top risk

8826

F12-NORTH · Water ingress

Indoor RH up 6w. 5 rain events in 30d. Prior leak — same riser.

Next move

7–14d

Inspect riser HW-12

  • Skip waterproofing
  • Save ~$140k
  • Trail logged

Cards reveal as you scroll — same way the engine builds the brief in the morning.

ROI

$0kAvoided / eventRestoration on a 1,000 sq ft incident.
~$0kSetup, 100k sq ftDOE EMIS · ~$0.05/sq ft.
~$0k / yrRecurringDOE EMIS · ~$0.02/sq ft/yr.
0%Median FDD savingsDOE Better Buildings.
Moisture damage in a wall corner with hairline crack.

1,000 sq ft event · published ranges

$23k–$45k

Mitigation $3–7.50/sf · Repair $20–37/sf.

Source · Palm Bld · DOE

Why now

BMS, sensors, meters, weather, build records — already there. We use what's there. No new hardware required.

$174.97BSmart building market · 2026~7%Median FDD savings~$0.05 / sq ftSetup cost~$0.02 / sq ft / yrRecurring~$3.01TFM market · 2026$23k–$45kPer leak event avoided

Architecture

Inputs → engine → daily score, ranked risks, next maintenance window. ML calibrates over time.

Sensors / BMSWeatherProject specsMaintenance historyNormalize → Zones / AssetsAsset + material modelRules + PhysicsEngineML LATERDaily score · 0–100Ranked risks + evidenceNext maintenance date
Isometric building data network with sensor nodes and connecting paths.

Wedge

Buildings · existing sensors · weather · build records. Then ML.

Not us

HVAC FDD · digital twins · embedded structural sensors. We sit above them — adding the maintenance-risk layer.

Pilots

Recruiting design partners. Three slots open.

Stage 01

1 building

Read existing data. Score in 2 weeks.

Weeks · not months

Stage 02

3 buildings

Calibrate. Tighten the FM loop.

Outcome feedback

Stage 03

Portfolio

Scale scoring + risk briefs across the book.

KPIs · prevented
Apply as a pilot

Owner-operators · developers · FM running sensor-rich commercial portfolios.

FAQ

  • No. We sit on top. Read what your BMS, sensors, meters already produce.

Team

A small team focused on this single problem. We read every reply and reach out personally.

Asjad Nirban portrait

Asjad Nirban

Founder

Master's in CS from University of Virginia.

Jeet Sharma portrait

Jeet Sharma

Co-founder

MS in CS from University of Massachusetts.

In validation · 2026

5 minutes of structured feedback. Or grab a Calendly.