Available today

Catch growth problems while the run can still be saved.

Around-the-clock monitoring of furnace power and temperature that flags events tied to bad outcomes, mid-run.

Built and run in production by our teamFirst useful result in a few months

The problem

Growth problems, from subtle to extreme, line up with historically bad outcomes, but they're caught only after the fact. Runs that could have been saved, or stopped early, are lost.

What we built

A monitoring system that watches power and temperature around the clock and flags events, subtle to extreme, that are highly correlated with historically polycrystalline outcomes.

What the monitor watches during a runIllustrative, synthetic data

Look closer. A subtle wander, easy to miss on a readout: worth a closer look while the run continues.

Act now. A clear event that calls for action now, while the run can still be saved or stopped early.

What changed

Growers got a mid-run early warning on events tied to bad outcomes, and outcomes improved sizeably over time.

How it works

  1. 01

    Connect to the furnace logs you already keep. No new sensors.

  2. 02

    Learn what your growers already suspect, and test it against what actually came out of the furnace.

  3. 03

    Flag two kinds of events: subtle ones that call for a closer look while the run continues, and clear ones that call for action now.

  4. 04

    Alert your team. People decide; the system never takes control of the furnace.

Catch what you'd miss

Subtle signs your growers wouldn't spot in time.

Act sooner

When it happens, not at the next log review.

Both save furnace time and material.

What carries over to your process

The telemetry pipeline and monitoring approach carry over to other materials and furnaces. The event models are rebuilt for each furnace.

What it takes

Your existing process data and some of an engineer's time. You set success criteria in writing before we start, and you keep every deliverable.

How a pilot works

Why not build it in-house?

The hard part isn't the alerting. It's merging years of logs from different controllers into one trustworthy record and proving which signals actually predict a bad boule. We've done that work in production across multiple furnace designs.

Tell us about one furnace.

What you grow, what you already measure, and where runs go wrong. No deck, no preparation, no commitment.

Talk to us about a pilot