Skip to content

Modules and situation assessment.

For specialists in IT, data protection and procurement: how MAKOR is built, which modules exist and how data becomes a verified assessment of a situation.

Structure

MAKOR separates infrastructure from application.

A shared infrastructure creates the foundation. Modules on top of it create concrete decision spaces. New modules need no new data integration.

InfrastructureÄther

Six building blocks form the shared layer above your IT.

Each building block can be used on its own. Together they carry all modules and your own applications.

01 · Connectivity

Connect instead of copy

Standardised adapters for GIS, ERP, logs, sensors, open sources and public authority interfaces. Data is referenced, not copied.

02 · Ontology

One shared language

A domain-specific vocabulary of actor, place, event and time translates between the source systems. Data classes and relationships can be modelled without code.

03 · InferenceCortex

Inference and models

Rule-based and ML methods with a model registry and a confidence score for each result. They run on your infrastructure, and no data leaves your scope.

04 · Harness & audit

The framework around the AI

The harness binds the AI to your data: it answers only on this basis, names sources and flags gaps. For every inference, the audit trail records which source, which model, which confidence and which point in time it is based on.

05 · Data flowsFlux

Orchestrated data flows

Declarative pipelines connect sources, models and results. Versionable, testable and repeatable, from pilot to regular operation.

Watch the film ↓
06 · Sovereignty

You remain the owner

Data, models and audit logs remain yours.

How data flows come about in MAKOR.

Flux
What the film is about

In two minutes the film shows how a data flow comes about in MAKOR: sources, models and results are connected in a declarative pipeline, versioned, tested and taken from pilot to regular operation.

Read transcriptSubtitles

Every team has them: small programs that fetch, transform and send data. And somewhere on some machine they run … hopefully. With Flux you simply create a folder with your script and upload it. Flux recognises it on its own. No form, no rebuild. You tell Flux when to start, and Flux takes care of the rest. Every run happens in its own self-contained capsule and brings along what it needs. Afterwards it is cleaned up. Nothing gets in anything else's way. It runs on the laptop just as it does in the data centre. Runs can trigger each other: the next step on success, plan B on failure. You see live what is running, what worked and what did not. Secrets stay secret. And if something goes wrong, Flux tells you. And if something stalls, Flux does not lose the thread. So what does this look like in practice? An example from the port. A new aerial image arrives, and Flux gets going. The AI looks for what you tell it in words. What cannot be is thrown out. What is new lands on the map as knowledge. And from there it carries on by itself. Flux is the pacemaker in Makor. It sets data, AI, knowledge and your own workflows in motion. Makor Flux. One script. One push. Runs.

Modules

Three modules for space, relationships and time. Your own applications complement them.

Module · SpaceAtlas

From pixel to situational picture.

Connects satellite imagery, maps, sensors and specialist applications into a spatial situational picture. Every statement carries its source and confidence.

Capabilities
Spatial data integration, image analysis and object detection, spatial analytics
Sources
Satellite, aerial imagery, sensors, vector maps, GIS, public authority data
Output
Situational picture tiles, risk maps, route analyses
Use cases
Detect changes along routes, map damage situations, plan land use
Module · RelationshipsSocius

From a single actor to the network picture.

Connects actors, organisations and events into a picture of relationships. Every connection carries its source and confidence, and a trail backs it up.

Capabilities
Actor and relationship detection, content and topic analysis, pattern recognition
Sources
Open sources, registers, master data, logs, event streams
Output
Relationship graphs, risk assessments, clusters
Use cases
Map dependencies in supply chains, screen business partners, link events across sources
Module · TimeChronos

From snapshot to trend.

Makes patterns over time visible: trends, anomalies and early indicators. Every forecast carries a confidence corridor and named assumptions.

Capabilities
Time series integration, anomaly and pattern detection, forecasting and early indicators
Sources
Sensor time series, logs, market data, event streams, public authority data
Output
Trends, anomaly markers, forecasts with corridor
Use cases
Load forecasts, telling sensor faults from real events, tracking the effect of measures
Your own applications · SDKCustom

Your team builds on the same infrastructure.

An SDK opens up the same infrastructure the modules run on. Your own applications inherit sovereignty, ontology, audit trail and explainability. No second stack, no second audit pipeline.

Building blocks
Connectors, pipelines, your own models, ontology extensions, UI components
Languages
Python, TypeScript, CLI
Use cases
Industry-specific situational pictures, integrating models from your own research, automating approval processes
Assessing situations

An assessment becomes robust when space, relationships and time come together.

Hypothesis networkSample data
  1. 01

    Collect observations

    From all connected sources, organised by place, relationship and point in time.

  2. 02

    Form hypotheses

    MAKOR looks for the best explanation of what has been observed and formulates possible explanations, each with a confidence score.

  3. 03

    Name gaps

    MAKOR indicates when the available data is not enough to confirm or reject a hypothesis.

  4. 04

    Verify

    Every hypothesis refers to its sources and its processing path. The audit trail records both.

  5. 05

    Decide

    Your specialists assess the situation and decide. MAKOR provides the basis.

Sovereignty

Sovereignty is the starting point, not an add-on.

Every component is built so that data, models and audit logs stay within your scope.

Location
Development, operation and hosting in Germany
Hosting
On-premise, private cloud or European PaaS, your choice
Operation
Air-gapped operation supported
Access
Bonn Systems has no access to customer data
Training
You train MAKOR with your data, on your infrastructure
Compliance
Compatible with the EU AI Act, BSI C5 and ISO 27001 audits
Foundation
Core infrastructure built on open source, no lock-in
Next step

Test MAKOR on your own question.

To the challenge →View industries