Home — About
About Choir Corp

We work on the assets where being wrong is expensive.

Choir Corp puts a written record behind every call on expensive power assets — data-center power equipment, battery storage and solid-state transformers. We tell you what’s going to fail, when, and why — while there’s still time to do something about it.

Who we are

We use whatever method actually works on your asset.

Choir is not an "all about physics" company. We predict how expensive assets degrade — when, why, and what it costs to wait.

Physics-informed models where the physics earns its place; statistical models where it does not. We have tested both across three asset classes, and the answer is not the same every time. What makes the models credible is the same thing that makes them honest: every number traces to a dated report, and when our own tests favour the simpler model, we say so.

Choir is the engine. The SENTRIX™ Alliance — Spearix sensing and LinkWorx private-carrier connectivity — moves trustworthy data to it from brownfield and secured sites that were never wired for it.


Leadership

Who you will be working with.

Diane Zuckerman

Diane Zuckerman

Chief Executive Officer
LinkedIn ↗
SH

Steve Harter

Chief Technology Officer
LinkedIn ↗
DY

David Yamaguchi

General Counsel
LinkedIn ↗

At a glance

One engine, whatever you own.

1
Prediction engine — the right technique chosen per asset class.
4
Products, first call to written record: PredictPowr™, FlowPowr™, UptimePowr™, GridPowr™.
3
Asset classes tested on public or simulated data.
3
Alliance partners — Choir, Spearix, LinkWorx — one deployable stack.

● The discipline

We publish our misses.

Across four independent evaluations, the evidence does not support a blanket claim that physics improves accuracy — so we do not make one. What we have is rarer and more valuable: a company that scores itself honestly, publishes the cases where a simpler model won, and can trace every number to a dated source.

See the evidence →
Published misses
MIT/Stanford real cells — plain model 1st, physics 4th
IGBT reproduction — a standard LSTM beat our physics model
NASA ablation — we corrected our own attribution

Reported as found.

Start with the data you already have.

A read-only assessment proves the forecast on your own data before any hardware conversation. You see the record; you decide what it is worth.

Software-first · no raw data leaves your network · every number traces to a dated report