Knowledge systems · Research

Knowledge grows.
Understanding connects.

A research direction for how knowledge is gathered,
connected, and carried forward.

Research in Progress

ARX

Research into the knowledge system intended to become the operational brain of Katanoesis.

ARX is an internal system being researched to support the company's work with connected knowledge. It is not presented as a commercial product.

Research in Progress

ARX-MMS

Knowledge management and exchange in support of ARX.

Outside workers bring knowledge in and receive knowledge back through ARX-MMS, supporting ARX’s own internal workers and managers. ARX-MMS is also being researched as a potential future commercial product, not a release commitment.

ARX · Internal knowledge system

ARX-MMS · Knowledge exchange

Knowledge takes shape.
A prompt brings it to life.

ARX + ARX-MMS · An interactive concept

Choose a prompt, as you would give one to a language model. ARX-MMS carries the request into ARX, where managers and workers retrieve what Katanoesis has established, and brings the answer back.

ARX Knowledge networkKnowledge at rest

Knowledge recordInternal workerArea manager

Response

Knowledge is already here.

What Katanoesis has established is stored, sourced, and connected. Choose a prompt to follow the request through ARX and back.

Illustrative network · 168 linked records

Conceptual simulation · Research in progress. The responses are accurate information about Katanoesis. The network, its workers, and the way requests travel are an illustration, not a working product or a map of how ARX is built.

Read the five prompts and responses

Explain how Katanoesis decides what to trust.

Nothing is trusted because it sounds convincing. Existing knowledge, assumptions, and proposed solutions are broken into their parts. Each part is tested against evidence, reality, counterexamples, and competing explanations. Only what survives that scrutiny is rebuilt, around the actual problem.

Sources only · Every claim stays attached to its evidence.

Show what the maturity labels mean.

Katanoesis separates six states: research in progress, exploratory work, in development, prototype, demonstrated, and released. Today, Euscholia is in development; ARX and ARX-MMS are research in progress. The rule: show what exists, explain what is being investigated, state uncertainty, and never present intentions as accomplishments.

Memory only · Each label is kept with the work it describes.

Map where Katanoesis works.

Where three concerns meet. Intelligence: reasoning, analysis, optimization, automation, and useful action. Knowledge: reliable information, memory, context, verification, and understanding. Protection: privacy, security, reliability, integrity, and human control. Katanoesis works at their intersection.

Connections, supported by Sources + Memory · Three concerns, one set of links.

Explain why Protect comes last.

Only in the list. The full method has seven steps: Observe, Understand, Deconstruct, Verify, Reconstruct, Apply, and Protect. The four public principles pair them. Protection is not added at the end; it constrains every step, preserving people, privacy, knowledge, security, reliability, and human control.

Context, supported by Sources + Memory + Connections · Protection reaches every area.

Describe how Katanoesis uses AI.

The method comes first; AI makes it faster. We observe and gather information, pick it apart until only its essence remains, and rebuild that essence to serve the goal, verifying and auditing at every stage. Then we apply it to reality; if it doesn’t hold, we run the process again. AI assists with research, criticism, verification, reconstruction, drafting, and auditing, but it is not the authority: interpretation, authorship, judgment, and publication stay human.

Context, supported by Sources · Tools assist; people decide.

Evidence before claims

Research earns
its conclusions.

We begin with questions, test assumptions, and share only what the evidence supports. Both systems are in early research. Experimental prototypes are part of that research: we test what works and what doesn't, analyze why, and refine what survives before anything moves into development. Public updates will follow verified progress.