Running it

Three hundred units. Nobody watching.

A fleet of services and scheduled jobs that keep each other honest. Each one reports its state; a supervisor notices silence; a dead-man switch rings when the supervisor itself goes quiet.

-units in the fleet
-working right now
-GPU utilisation

Principles

01

Verify before you work

Every session begins with one call that checks servers, services, the model, the database and the vector store. Numbers are read live, never remembered.

02

Restore beats backup

Backup integrity and the ability to restore rank directly after data integrity. A backup nobody can restore is a hope.

03

Silence is an alarm

Heartbeats and a dead-man switch: a stopped watcher is itself an event.

04

Change with a way back

Every production change starts with a copy and a one-line rollback.

05

One owner per concern

Services have a single purpose and a written header: version, date, author, location, purpose.

06

Pay once

Local models answer first. An answer bought from an outside model is stored, so it is never paid for twice.

Priority, in order

Free access

See the live system.

Register once, free, and the architecture and live demo open for 30 days. Payment is only for the reference code.

Register — free →
Behind the door

Read how it is built.

Six reference modules in textbook form: access, telemetry, payment, self-improvement, grounded retrieval, consent-first measurement.

Open the door — EUR 100 incl. VAT →
RiNET · Engineered in the Netherlands · Operated on sovereign European infrastructure · Privacy