OWNERSHIP / GOVERNANCE

Ownership is not a feeling. It is a list.

What we build with you sits in your environment, in your name, with the documentation and operating rhythm your team needs to keep running it.

01 / WHAT YOU HOLD

The ownership checklist.

Seven things your team holds after the build. Each one is a real artefact in your environment, not a setting in someone else's account.

The ownership checklist Seven items the team holds after the build, each shown as a ticked row: the data, the signals, the model, the decisions, the agents and workflows, the measurement, and the capability. Each is a real artefact in your environment, not a setting in someone else's account. WHAT YOUR TEAM HOLDS The data customer_data.layer The signals customer_signals.set The model model_logic.spec The decisions decision_rules.map The agents and workflows workflows.ops The measurement measurement.board The capability capability.runbook
Seven things your team holds after the build. Each one is a real artefact in your environment.
customer_data.layer

The data

Customer records, source mappings, pipelines, transformations, and rules inside your environment.

customer_signals.set

The signals

The behavioural, commercial, lifecycle, and service signals used to understand customers.

model_logic.spec

The model

Documented and versioned logic for scoring, segmentation, prediction, prioritisation, or recommendation, written so your team can change it.

decision_rules.map

The decisions

Rules, triggers, next-best actions, and decision histories traceable to the data behind them.

workflows.ops

The agents and workflows

Lifecycle, CRM, sales, and service workflows, with the AI agents working inside them.

measurement.board

The measurement

Dashboards, alerts, outcome reporting, and improvement backlog.

capability.runbook

The capability

Runbooks, documentation, training, and the operating cadence.

02 / OWNED, NOT RENTED

Same intelligence. Opposite outcome.

The record can be owned or rented. So can the AI. The difference is whether your team can see it, run it, and keep it.

Same data and the same AI, owned or rented The same data and the same AI split into two territories. In your environment they are sealed and owned: you can inspect the logic, run it yourself, and keep it after exit. In a vendor account they are dashed and metered: you pay to query, the logic is hidden, and it leaves with the contract. SAME DATA. SAME AI. data YOUR ENVIRONMENT Inspect the logic Run it yourself Keep it after exit OWNED. NO RENT. VENDOR ACCOUNT Pay to query Logic stays hidden Leaves with the contract PAY TO ACCESS
The same customer data. Owned in your environment, where your team can see it, run it, and keep it; or rented from an account you pay to enter.
OWNED RENTED
The record is in your environment.
The record sits in a vendor account.
Logic is documented.
Logic is hidden in settings.
Model changes are versioned.
Model changes are not fully visible.
Workflow rules can be inspected.
Workflow rules depend on supplier access.
AI works in your systems and leaves a record.
AI sits in seats and leaves nothing behind.
Reporting ties decisions to outcomes.
Reporting is limited to what the platform exposes.
Your team can run it.
Your team pays to keep querying it.
The supplier can leave.
The intelligence leaves with the contract.

Removability is the test. Remove NTWRK at a phase boundary and the completed work keeps running. We build to that line.

03 / GOVERNANCE

Built for explanation, not dependence.

An operating principle, not legal advice. When the records, rules, model logic, and decisions sit in your environment, the explanation of what happened and why is already written, in your name. From December 2026, many Australian businesses must state in their privacy policies the kinds of automated decisions they make. The law asks for the disclosure; owning the record behind it is the difference between showing your working and asking a supplier for it.

The record writes itself Three owned inputs on the left, the data, the rules, and the decisions, flow into a single produced artefact on the right: the explanation of what happened and why, already written in your name. It is tagged to the December 2026 requirement that businesses explain how their automated decisions are made. IN YOUR ENVIRONMENT data rules decisions THE RECORD EXPLAIN THE DECISION / DEC 2026
When the data, rules, and decisions sit in your environment, the explanation of what happened and why is already written, in your name.
data_lineage.note

Data lineage note

decision_rules.summary

Decision-rule summary

model_change.history

Model-change history

workflow.map

Workflow map

access_responsibility.map

Access and responsibility map

handover.runbook

Handover runbook

QUIET INTELLIGENCE BEATS LOUD PLATFORMS

See whether building it to keep is right for you.

One conversation about what you rent today and what your team should still hold after the work is done.