Your know-how
Documents, experience and rules we are allowed to use.
Our approach
Understand first.
Then choose the right tools.
Our interdisciplinary bigRing methodology connects domain knowledge in a model of roles, procedures, relationships and outcomes. Rules and exceptions come with their reasons and the context in which they apply. Memory and individual tools play different roles in this approach.
bigRing in plain language
A shared language
for your working reality.
A model is more than a picture. It clarifies who does what, under which rules and conditions, and with what expected outcome. We describe causes and consequences according to the available evidence; the shape of a graph does not establish them.
Rules, exceptions and their reasons remain in a shared context. For important claims, we distinguish verified knowledge, assumptions and open questions, keeping their sources traceable.
Illustrative model of one area of work; the map can be scrolledbigRing
Sources
What we rely on and where knowledge comes from.
People and roles
Who acts, decides and verifies.
Processes
What happens, in what order and under which conditions.
Relationships and agreements
What connects participants and defines their commitments.
Objects and outputs
What people work with and what should be produced.
A trustworthy foundation
We distinguish
what we are relying on.
An expert-approved rule, practical experience and a working hypothesis are not the same. The model needs to distinguish their sources, validity and limits.
- A source for important claims
- Validity and version of a rule
- Marked uncertainty
- A person responsible for content
Intelligencer / CausalMemory
Knowledge has a place.
And a history.
Intelligencer is an environment for managing and using knowledge. CausalMemory supports memory, versions, data provenance and retrieval. We verify deployment options and available interfaces for your task.
bigRing
Shared terms, relationships and agreed meaning.
Intelligencer
Management, versions, source provenance and retrieval.
People and agents
A specific task in a selected environment.
Technical details for your IT team
CausalMemory source material describes PostgreSQL with versioned nodes and edges, temporal validity and provenance. We verify the specific interface, API or MCP connection, configuration and readiness for your use. This description does not confirm deployment of every capability.
| Layer | Purpose and boundaries |
|---|---|
| Knowledge model | bigRing expresses terms, relationships and patterns. Graph storage alone does not establish truth or causality. |
| Memory and retrieval | CausalMemory retains versions and provenance, combining relational, graph and vector methods according to the implementation. |
| Advanced layers | The extended cognitive model and Apache AGE require separate validation. Their presence in a development branch does not establish deployment. |
| Agent runtime | Agno or another agreed tool may be used. Memory and rules should be separated from the specific runtime to the agreed extent. |
| Operations | Hosting, identities, isolation, backups, recovery and limits are confirmed for the specific use and contractual scope. |
Your environment
Usable where
you work.
A web interface, Microsoft 365, another working system or an agent connected through API or MCP. We choose according to need, permissions and reasonable costs.
- First, we assess the options in your current tools.
- The integration scope is confirmed before commissioning.
- Permissions remain separate from memory content.
- Export and portability are agreed in specific formats.
Boundaries for responsible use
A good foundation needs
good verification too.
Organised knowledge can improve AI use, but does not remove every error by itself. We test sources, context selection, permissions and outputs against agreed scenarios.
Illustration: a new service colleague needs to know whether they can approve a repair under warranty. Instead of hunting for one sentence in documents, they get the rule, its conditions, an exception and its reason, the source and the person who decides when something is unclear. A graph alone does not prove causality, and memory does not guarantee error-free AI.
First step
Do you have your own technical
environment or agent?
Let’s explore how to prepare the right knowledge for it. You do not have to assume a full system replacement first.
Discuss one area of work
First, we clarify your needs and a suitable scope.