Mohamed Yassine Hemissi

Student • Dreamer • Human

HomeProjectsBlog

© 2026 Mohamed Yassine Hemissi. All rights reserved.

← Back to projects

Notion2Pi - Abstract Math Visualized

Flagship

An AI-assisted web app that turns abstract mathematical notation into more intuitive, multi-perspective learning experiences through a 7-vector analysis framework.

2026·By Mohamed Yassine Hemissi·Updated January 5, 2026
EngineeringCreative LabAcademic/Community

Technologies Used

Next.jsTypeScriptPrismaTursoZodAI Orchestration
View on GitHub →Live Demo →

Notion2Pi Demo

Notion2Pi is a web application designed to make abstract mathematics feel more intuitive and accessible. The idea was to move beyond dense symbolic notation and turn formulas into something a learner can approach from multiple angles rather than as a single static definition.

Core Idea

At the center of the project is a 7-vector analysis framework that looks at a formula through seven dimensions:

  • role
  • domain
  • binding
  • variance
  • geometric view
  • invariant properties
  • limiting behavior

Instead of only presenting a formula as text, the system reframes it as a conceptual object that can be explored, interpreted, and visualized.

What It Does

The platform includes a formula catalog, dynamic formula pages, and an AI-backed generation layer that helps turn mathematical content into richer educational material.

The goal was not just to display LaTeX nicely, but to help users understand why a formula exists, what its parts are doing, and how it behaves under change.

Architecture

The project was structured as a modern Next.js application with clear separation between UI, business logic, validation, and database concerns. It also includes a flexible orchestration layer for working with multiple AI providers.

Why It Matters

What makes Notion2Pi meaningful to me is that it sits at the intersection of mathematics, pedagogy, and software design. It treats mathematical notation not as something to simplify away, but as something to reinterpret into forms that are easier to reason about.

Repository

The implementation details and full README are available on GitHub.