AETHOS produces a decision object for one estate: an asset model, the sources that attach to it, and a ratified policy that says what outranks what. All three are specific to the estate, and all three have to be built with the people who run it. That build is what this engagement is.
A license key and a manual is not a deployment. Praelor engineers and practitioners work alongside your team through each of these, and each one has a written output your side keeps.
An inventory of what exists: sensors, security monitoring, control systems, field surfaces, reporting lines and the organizational structure that already routes them. Written before a single line of configuration.
The estate written down as assets with owners, so that a signal has something to attach to and two signals on the same asset can be recognized as one story rather than two tickets.
Each source connected as an adapter against a common contract, normalized and validated against traffic your team recognizes. Field surfaces such as TAK Server, ATAK and WinTAK, and control system sources such as OPC UA, Modbus and DNP3, where they are in use.
The weights, thresholds and windows that decide what outranks what, set in a working session and signed by a named person in your organization rather than shipped as a vendor default.
The same decision object shaped for each altitude that has to act on it, from the post or the floor, through the watch officer, to the executive who answers for posture.
The queue run against real traffic and compared item by item with the judgment of the people who already do this work, against acceptance criteria written before the run rather than after it.
The sequence matters more than the speed. A model built after the sources are connected is a model nobody trusts, and a policy inherited from a default is a policy nobody will defend.
We inventory the sources that exist, the ones that are specified but not built, and the ones that are neither. All three are recorded. Nothing is configured before this is written down.
The estate becomes objects with owners, using your own naming rather than ours, so that the queue that follows is readable by the people who already work this ground.
Sources are added as adapters against a common contract, one at a time, each validated against traffic your team recognizes. What does not connect is recorded on the coverage statement instead of being absorbed quietly.
Weights, thresholds and windows are set in a working session with the people who hold the authority, and signed. This is a decision your organization makes and can revisit, not a default we ship.
The object is shaped for each altitude: what changed on this shift at the front line, the ranked picture and option set in the operations center, posture and its basis at the executive desk.
The queue runs against real traffic in parallel with current practice, and is compared against what your people would have done. Disagreements are worked back into the policy, not explained away.
Configuration records, adapter specifications and the coverage statement pass to your team, with a defined support period before steady state so that ownership actually transfers.
Nothing here assumes a complete estate or a rip and replace. Sources are added as adapters over time, and the score states the coverage it was computed on.
Where a feed does not exist, the gap is disclosed on the object.
A missing source is a fact about the estate, and the decision object carries it rather than absorbing it into a number that reads as complete. That is also how the deployment stays honest as new sources are added later by your own team.
Hosting, network placement and any restriction on connectivity are settled during the assessment phase against your requirements. We state what is built and what would have to be built for a given arrangement before anything is agreed.
We walk your source inventory and name what is built, what is specified, and what is neither.