While the machine is still being built

Millwright vs Sift

Sift stores high-rate test telemetry and lays this run over the last one, so a campaign stops depending on one engineer scrolling. Millwright is for after the machine ships: the historian, the alarms, and the work order are already there, and the question is on shift.

Keep Sift when: You are bringing a machine up, and the question is run-over-run.
Pick Millwright when: The machine is in service, and the question is why it tripped and who closes it.

Millwright Trace, plant telemetry on an asset. — Millwright

Sift reviewing a chamber-pressure anomaly across a flight run. — Sift

The difference

The question

Millwright: Why this unit tripped, and who is closing it.
Sift: What changed between this run and the last one.

When

Millwright: The machine is in service. A crew runs it.
Sift: The machine is being designed, tested, and brought up.

The data

Millwright: The historian, the alarms, and the work the plant already has.
Sift: High-rate, high-cardinality telemetry from stands, vehicles, and campaigns.

Review

Millwright: Diag shows the points it read. A person approves the close.
Sift: Automated review across a campaign's channels.

Who

Millwright: Operators, planners, contractors, stakeholders.
Sift: The hardware engineers on the program.

Where it sits

Millwright: On the historian you have. Does not write to the DCS.
Sift: Its own store. Sift integrates with Grafana rather than replacing it.

Feature comparison between Millwright and Sift
Feature	Millwright	Sift
The question	Why this unit tripped, and who is closing it.	What changed between this run and the last one.
When	The machine is in service. A crew runs it.	The machine is being designed, tested, and brought up.
The data	The historian, the alarms, and the work the plant already has.	High-rate, high-cardinality telemetry from stands, vehicles, and campaigns.
Review	Diag shows the points it read. A person approves the close.	Automated review across a campaign's channels.
Who	Operators, planners, contractors, stakeholders.	The hardware engineers on the program.
Where it sits	On the historian you have. Does not write to the DCS.	Its own store. Sift integrates with Grafana rather than replacing it.

Sift as described on their public pages in October 2026. Corrections.

Where Sift wins

You are designing, testing, or bringing the machine up.
The data is campaign telemetry, not historian history.
Run-over-run review is the job.

Where Millwright wins

The machine is in service, and a crew runs it.
The historian and the CMMS already exist.
Scope must be smaller than the whole program.

How they fit together

Building the machine and running it are two years of the same asset. Sift follows the program from prototype into the operations of what it built; a plant that has run for decades on a historian is a different customer. A company that builds on Sift and operates a fleet buys both. The engineering room keeps Sift, and the operating company runs Millwright.

Questions

Sift talks about operations too. Does the line hold?

For the program that built the machine, yes. A constellation run by its own engineering team is their customer. A plant on a historian and a CMMS is a different room.

Should engineering stay on Sift after the fleet ships?

For what changed and which channel, yes. The crew, the planner, and the customer ask different questions, on different systems.

Does Millwright ingest high-rate test telemetry?

No. It reads the plant historian, live and recorded. A hot-fire campaign belongs in Sift.

Other comparisons

See it running on your own plant.

We are onboarding a small number of operations teams. Three questions, and we will be in touch.