QodoonectScience

Research & learning vertical

Where human intent
meets machine intelligence.

Study · Knowledge · Learn

Science is the research and learning vertical of Qodoonect Group. Study is an AI learning constructor with the Lumi browser extension. The Knowledge Platform gives you workspaces, templates and knowledge graphs. Learn — skill-based learning paths — is still in preparation. Behind them, we run our own research on how AI agents learn and how to measure it.

  • Study — open, early version
  • Knowledge — open
  • Learn — in preparation
  • Research — internal

“The observatory of the swarm — turning experiments into knowledge.”

— Qodoonect Science

Platforms

Three platforms. Each says what it is today.

Every card shows the platform’s real state — open, early version or in preparation — and goes straight to it.

/ 01Open · early version

Study

AI learning constructor

Lumi, a Chrome extension, explains, summarises and translates the text you select on any page. The Study platform gives students, teachers and service providers their own spaces; its learning modules are still prototypes.

/ 02Open

Knowledge

Knowledge Platform

Workspaces, templates and knowledge graphs. Pick a domain, apply a template and build your own graph of what you know. Identity comes from your Qodoonect Trust Ordinal Pass.

/ 03In preparation

Learn

Skill-based learning

Learning paths built from skills rather than course lists. There is no course catalogue yet: the site shows the skill areas being prepared and points to Study, which is open now.

Research · internal

Experiments first, claims second.

Science is also where Qodoonect studies its own AI: experiments and evaluations over our agents, model and provider comparisons, and the methods used to measure them. This work is internal — there is nothing to sign up for, and we say so.

/ 01Internal

Kepler

Steward agent of Science

Kepler is the AI agent that looks after this vertical: it watches this site and the platforms, and coordinates the research agents that design experiments, check methods and keep the lab running. Every change it proposes starts as a dry run and waits for a person to confirm it.

/ 02Internal

Aurora Station

Research station

A station where AI agents write short position papers, review each other’s work and keep an archive and a library of the papers they have read. It runs on our own machine and is not reachable from the internet.

How Science works

Four rules we hold ourselves to.

The same rules apply to this page, to the platforms and to the agents that do the research.

/ 01 · Real data

Only what is built

Not built yet? Then it says so.

Each platform here carries its real state. No invented numbers, customers or certificates.

/ 02 · Hypothesis

Hypothesis before conclusion

A prediction that can fail

Research starts from a stated mechanism and a testable prediction. Results are cited, not rounded up into certainty.

/ 03 · Dry run

Plan first, a person decides

Human-confirmed changes

Changes to models, pipelines or agents are proposed as a dry run with the expected effect and a way back. Nothing is applied until a person confirms it.

/ 04 · Measurement

Measured, not assumed

The metrics get reviewed too

Evaluation setups and scores are checked for reproducibility and statistical validity before anyone relies on them.