☰
KOEN

Personalized AI / A human-centered direction

Personalized AI begins with
understanding your brain.

We are designing a way for external AI services to use brain insights you choose to share when making their first suggestion.

Explore the possibilities ↗
PATENT FILING IN PROGRESS · IN DEVELOPMENT
Context delivery in design
You.AT THE CENTER
Your preferencesYour goalsYour present momentYour brain profile
CONCEPTUAL VISUAL · NOT A PERSONAL ASSESSMENT

Before the first lesson, a tutor could already prepare for
what comes easily to you and where a little help might matter.

01 / From people in general to you in particular

What a questionnaire
can miss.

A questionnaire can capture what matters to you. But describing how you usually learn or make decisions asks you to recall past situations and recognize patterns in them. Some patterns are hard to notice or put into words. BAS 2.0 adds a source of evidence that does not depend on that recall. It uses MRI measurements to build a brain profile in the Meta Quotient Report (MQR).

  1. 01
    Interpret the measurements.

    The starting point is a reviewed account of your strengths and potential difficulties, including uncertainty in each finding.

  2. 02
    Plan for the task.

    The assistant connects a relevant finding to a decision, such as when to revisit a new skill or how to guide a transition.

  3. 03
    Follow your response.

    What you do independently, where you need help and what you tell the assistant guide its next adjustment.

02 / The Neurogazer approach

From MRI measurements
to a first decision.

BAS 2.0 and the MQR give Neurogazer a foundation for interpreting brain measurements. We are developing an Individual Context Layer to carry a relevant, reviewed finding into a specific decision an AI assistant can make with your permission.

01 / MEASURE

Start with BAS 2.0.

BAS 2.0 analyzes MRI measurements. They offer a starting point even when a person has no chat history and finds a pattern difficult to describe.

02 / INTERPRET

Read the MQR in context.

Two findings can matter together. A learner’s profile may suggest both strong mathematical learning and difficulty holding several steps in mind. The interpretation preserves both.

03 / APPLY AND TEST

Choose what to prepare.

For that learner, the proposed tutor would keep the algebra challenging and build a short recall check into lesson one. The learner’s performance would guide the timing of the next review.

Explore the BAS 2.0 foundation ↗

03 / Illustrative scenarios

What changes in
the first suggestion?

There is an ease in talking with a close friend or family member who remembers what matters to you. These fictional scenarios imagine how an assistant could earn a little of that familiarity, starting with an MRI-based brain profile and remembering the experiences you choose to share.

Mira’s personal math tutor

Challenging algebra, with review timed for Mira.

Mira’s reviewed MQR profile suggests strong Mathematical Learning and a relative weakness in working memory. Before their first lesson, her tutor prepares a challenging equation and a brief recall exercise. Mira can show what she understands without first having to explain which part of learning algebra feels hard.

During the lesson
Mira explains why an equation stays balanced and solves a new one on her own. Ten minutes later she still explains the rule, but skips a step when the worked example is covered. Her tutor records exactly where she needed a cue.
At the next lesson
The tutor keeps the challenging equations and brings the first review forward to five minutes after the new example. Mira reconstructs the missed step from memory in a fresh problem, then returns to the harder work.
Setting her own review interval
Five minutes is a starting point. When Mira retrieves the step unaided, her tutor lengthens the interval and checks again the next day. If the same step slips, it brings that step back sooner. Mira’s unaided recall sets the pace for each skill.

05 / Responsible progress

What does brain analysis
add to support?

For each use, we need to establish whether a reviewed MQR finding helps an assistant choose a more useful starting point. In education, one test would be whether the first lesson offers an appropriate challenge with fewer adjustments.

Validate what adds value.

The additional value of MRI-derived evidence must be tested for each intended use, against support informed by questionnaires, stated preferences and everyday context alone. The proposal does not establish improved learning, care outcomes or personal prediction accuracy.

Develop with clear boundaries.

This is a development-stage concept. It is not presented as an available clinical service, a diagnostic tool or a way to determine a person’s future. People retain decision-making authority, with professional judgment where required. In the future, a robot could use this same understanding of a person to arrange a physical workspace around how they think.

Patent filing in progress. The technology is under development. Filing activity does not imply a granted patent, validated efficacy or a commercial launch date.

Decode the Brain. Elevate the Human.

From understanding
the brain to helping
the individual.

We welcome conversations with research, education, care and AI partners exploring responsible personalization.

Explore a collaboration ↗