Data pipeline builds

Fire is the first product. The same method applies to satellite or sensor data that has to reach a decision maker in minutes, such as a fire or emergency agency or a government team running a response: sources taken in as ground stations receive them, located, fused, validated, and delivered as data and alerts.

Requirements before any data moves

Data jurisdiction, processing, and availability requirements for each link in the chain: acquisition, transmission, storage, processing, and insights. Sovereign compute by default.

Built, then operated

We assign the sovereign compute each pipeline runs on and operate it for the customer. Availability is a requirement agreed at the start, and the service clock measures processing time for as long as we run it.

A pipeline measured against a service clock

Processing time runs from data arrival to data and alert sent. Accuracy is reported as error in metres, detection rate, false alarm rate, and F1 score against a reference source.

Labelled data for AI and simulation

Every record carries its location, time, and measured properties, derived from physics and computational analytics, ready for training AI models and for simulation.

Each pipeline is built to a control baseline agreed at the start, such as ISO 27001 or the Essential Eight, by people who have run cyber functions in regulated banking, payments, and insurance.

When to commission a pipeline

Satellite or sensor data arrives too late, or too coarse, to act on.
Several sources describe the same event and nobody fuses them.
Data has to stay in one jurisdiction from acquisition to insight.
An AI model needs labelled ground truth, built on physics.
Government or an agency must reach the data for as long as an emergency runs.
No team in house can run the pipeline around the clock through a season.

Compliance services

A fixed scope, fixed fee space cyber assessment for Australian space companies. Your prime or customer has asked a security question you cannot answer yet. In four to six weeks you get a report you can hand back to whoever asked, and a list of what to fix first.

A report written for the person who asked

The same document goes into a prime's security annex, a procurement questionnaire, or a board paper. Two assessors sign it after peer review, then walk it through with whoever has to defend it.

A fix list in order

Every finding carries a consequence rating from U1 to U5. Breaches you cannot undo come first, then design freeze items while the freeze is still open.

The evidence behind the number you sign

Under CMMC Level 2 you post your own score and affirm it every year. The report is what sits behind it.

The work is sized to what you fly. Payload class, bus model, and orbital layer set the scope, and ground segment only is a valid answer.

Assessment timeline and client responsibilities

The fee is fixed before we start, with no variation without your sign off. Your engineers spend about 12 to 14 hours across the four to six weeks.

Week 1

You send the evidence we list. Your team spends 4 to 6 hours this week and about 2 hours a week after that.
Weeks 2 to 4

We assess the evidence. Where it is thin, we run a working session before writing the finding.
Weeks 5 to 6

Peer review, then the report and a walkthrough with whoever has to defend it.

Assessment scope of work

One control set answers three regimes.

When to commission an assessment

A customer or prime has asked a security question you cannot answer yet.
Design freeze, while a control still costs a decision instead of a redesign.
Pre-launch acceptance.
First flight, and annually after it.

Start with a 20 minute call.

Enough to say whether a pipeline build or an assessment answers your question, what the scope looks like, and the fee.

Start a conversation

Services