Ruka hadi kwenye maudhui
Book · First Edition, 2026

Engineering Educational Intelligence

What a learner record must be — and what a system is entitled to claim it knows.

By Dennis Kariuki · Founder, EduNexus

Download the Book (PDF)

Free · No signup required · ISBN 978-9914-29-126-1

Why I wrote this

Two students sit for the same exam and score 65%. One climbed from 40%. One fell from 90%. The average is the same number for both of them — and it tells you nothing about which is which.

That's not a small rounding problem. It's the reason most systems that claim to understand a learner can't actually distinguish a struggling student from a recovering one. This book works through what a system has to get right before any AI can responsibly help a teacher — starting with the representation underneath it, long before a model ever enters the picture.

It's the argument behind what EduNexus is actually built on: evidence instead of grades, trajectory instead of averages, and an honest signal of how much of any of it is actually known versus guessed.

Inside the book

1

Learning Is Not Data

Why a single grade destroys the exact information a teacher needs — and what a record has to keep instead.

2

The Architecture of Educational Intelligence

The bounded-context architecture a system needs to reason from evidence rather than scores.

3

The Reasoning Engine

How reasoning holds competing explanations open instead of collapsing to one confident verdict.

4

Computational Intelligence

Where a language model is structurally entitled to help — and where it explicitly is not.

5

Operational Architecture

What it takes for all of this to keep working correctly, for every learner, every day, including when something fails.

6

The Institution

What actually changes for a school, a teacher, and a parent once the record is honest.

About the author

Dennis Kariuki is a Kenyan educator, software engineer, and founder of EduNexus, where he works on educational intelligence systems designed around evidence, learner progression, and human-guided decision-making. His work sits at the intersection of classroom practice, educational systems design, and software architecture.

Online: 8am - 8pm