Control over where the data lives.

No invisible copies, a named operator, and an exit you can describe.

Deployment, residency and operational ownership are design decisions, not add-ons. This page states what you can rely on today, and what the platform does not promise before it can be demonstrated.

The problem

Every copy is defensible. Nobody decided the total.

No breach is required for data to cross a boundary: ordinary copies do it, each made for an ordinary reason. The sovereignty problem is rarely dramatic — it is arithmetic.

The copies

A cache for speed, a pipeline for reporting, an export for a model, a backup for safety — four reasonable copies, four new places the data can be reached, and no list that keeps up with them.

The distance

The system is bought from one provider, run by another team, fed by pipelines a third built. Nothing malicious in the arrangement — just distance, and distance is why reach stops having a single answer.

The edges

The perimeter is drawn around the primary store; the export, the cache, the nightly extract and the vendor’s copy sit outside it. Control that ends where the copies begin is not control — it is a line on a diagram.

The question

Three questions every infrastructure decision has to answer.

They are boring questions, and they are the ones that decide whether a platform can be adopted at all.

Residency

For some organisations the location of data is a legal requirement, not a preference. That constraint has to be designed for, not bolted on.

Ownership

Someone has to operate the system at three in the morning. Being able to point at who, and where, is part of the product.

Exit

A data layer you cannot leave is not infrastructure, it is a dependency. The shape of that exit is decided by the storage contract, not by a marketing page.

Data control

One layer between your data and everywhere it could live.

Location, control, replication and access are the four questions the visual answers. The destinations it points at are direction, and are drawn that way.

your data rows · documents
events · vectors
PLOMID one storage contract

location · control · replication · access

On-prem The same layer placed on hardware you control, in a location you choose, with residency governed as declared policy. This is what the deployment fabric is designed around — the fabric is not available yet.

Control, dimension by dimension

Seven questions, answered honestly.

Seven questions every control story has to answer, stated plainly — including the parts that are still being designed.

Data location

There is one system and one place your data lives, so there is no ambiguity about which copy is authoritative.

Data residency

Region and residency boundaries as declared policy, part of the deployment fabric being designed.

Ownership / control

Who operates the system, and where, is part of the first conversation rather than a plan tier on a pricing page.

Access

One system to reach, with access decided between you and the people running it, there is no separate fleet of stores to reason about.

Replication

Where copies may exist, how many, and what keeps them consistent. This is a designed property, not a side effect of setup.

Jurisdiction

Which infrastructure and which data may fall under which rules. Tied directly to residency, and equally unfinished.

Portability / exit

One storage contract means the shape of the exit is a property of the design rather than an export tool bolted on afterwards.

The words

Four words, used precisely.

They are not synonyms, and most deployment conversations turn on the differences. Each word is used in one way on this page.

  • Sovereignty Operational control: where the system runs, who operates it, and what can be shown about it. A property of the deployment, not a claim on a page.
  • Residency Where the data is. A constraint the deployment has to satisfy, and part of the fabric still being designed.
  • Jurisdiction Which rules apply to which infrastructure and which data. It follows residency: change the location and the rules change with it.
  • Portability The ability to leave with the data, and to be able to describe how. A property of the storage contract, not a tool bolted on afterwards.
In the product

What you can rely on now.

Five things that are true today, not aspirations. The parts still being designed are stated where they appear.

One system, not a fleet

PLOMID runs as one system with one storage contract. Fewer components means fewer places where a copy of your data can quietly exist.

Runs on your hardware today

The Community edition installs on your own hardware — one-line shell install, the Docker image, or the source — and every tier runs the same layer. What is still being built is the deployment fabric that turns placement and residency into declared policy.

No account to try the surface

The playground runs entirely in your browser against a dataset that ships with this site. Nothing you type there is sent anywhere.

This site profiles no one

No advertising, no cross-site tracking, no cookies set by the site itself. Aggregate visit measurement runs through Google Analytics; the playground stays entirely in your browser.

Access is a conversation

Where the system runs, and under whose operations, is part of the first conversation rather than a plan tier on a pricing page.

Why it matters

Sovereignty is a decision at design time, a project at every other time.

Permissions can be revoked in a second; a copy cannot be unmade. That asymmetry is why the questions arrive on their own schedule — answered by the design, or reconstructed under pressure.

The review

Security questionnaires and diligence processes ask the same three things: where the data lives, who can reach it, and how you would leave. The answer is either a pointer to the design or a project with a deadline.

The incident

Under pressure, the useful map is the true one: where the data actually sits, who can actually reach it. One storage contract has one answer to give; a system of copies has an inventory problem at the worst hour.

The change

A new region, a different operator, a migration: the same questions again, on a timetable you do not set. Every change is priced by the first design.

The conversation

From a constraint to a shape.

Three steps, the same each time, whichever constraint you bring. Deployment and control are settled in conversation before they are settled in configuration.

State the constraint

A residency boundary, an operator requirement, or an exit you have to be able to describe. The sharper the constraint, the more specific the answer can be.

Decide the shape together

Where it runs, who operates it, and what has to be provable. These are decisions made with the team, case by case, rather than selected from a tier list.

Write the answer down

What is true today, what is being designed, and what is direction — in the same language this page uses, and nothing described as available before it can be demonstrated.

Questions

Deployment and control.

Deployment and control

Where it runs is a decision,
not a default.

Where the system has to run, who operates it and what has to be provable are decided with the team, case by case. Describe the constraint and the answer is specific to it.