Deployment

Deploy the protection where you need the data to be processed.

Where it runs shapes the data flows, the administration and the cost. We go through what should run at a provider and what in your own infrastructure, and propose a deployment for your actual use.

Operating environments

Three environments, one protection

A comparison of environments, not price levels. A dedicated server is an optional deployment of the same Corpilus, never a higher security package.

Operated by the provider in the EUOperated by the provider in the EU

You start without installing anything — we run the agreed solution in data centres inside the European Union.

The proposal names the services, the place of processing, who has access and which responsibilities are whose.

Dedicated instanceDedicated instance

An environment set aside for your company alone, by agreement.

The proposal names the location, the isolation, the capacity and how the instance is administered.

Your own infrastructureYour own infrastructure

The agreed parts of the protection run in your environment.

The proposal names the supported components, how they are updated and which external dependencies remain.

Hardware option

NVIDIA DGX Spark is an example of hardware for a local deployment that suits one — not a fourth product. It does not follow that every protection needs a GPU or a local language model.

Before the proposal

The whole data flow decides.

A server of your own does not by itself mean that no part of the solution communicates outwards. We go through the models, the checks, the external services, the updates and the administration together, and state the conditions in the deployment proposal.

  • Where the content is processed
  • Who has access to it
  • Who administers the operation
  • What you need to prepare

For the deployment itself we draw the components up one by one: input, output, location, retention, dependency and responsibility. Those values belong to the deployment register and are confirmed there for your deployment, not carried over from this page.

How it starts

Five steps to a working deployment

  1. 01
    Pick one flow
    One channel where the data actually passes, not the whole company at once.
  2. 02
    Check the environment
    What runs where today, what it depends on and who administers it.
  3. 03
    Agree the scope
    Which channels are connected, what is recorded and where a decision may be carried out.
  4. 04
    Connect and verify
    We go through the acceptance scenarios together, including the behaviour when a check does not run.
  5. 05
    Hand over a report
    What was connected, what it found and what is left open. Measurement can be part of it; it is not a universal obligation.

The licence, the integration, the infrastructure and the operation can be separate costs. You get the overview in the proposal. How the price is put together

A local model is not free operation: a provider's tokens are one cost, and the hardware, licence, energy and administration are others.

Next step

Let's find the deployment that fits your conditions.

Tell us where the data has to be processed and who administers the environment. We will propose a deployment and name its conditions.