Cultivate
Case study · AI Tools & Strategy
A connectivist learning environment: a personal hub for professional development and a curated news feed. The first AI build here, the experiment that started all of this.

Goal
Keep AI and EdTech professional development organized in one place, and surface the right resource, deadline, or reading at the right time. It replaces scattered bookmarks, notes, and calendar reminders with one calm space designed around real working conditions, between classes, in office hours, or offline on a plane.
Cultivate is a personal learning environment built on connectivism, where the network of people, tools, and sources is the point rather than a single closed app. It is a personal hub for professional development plus a curated feed of AI-in-education news and futures reading, trainings, and conferences. It is a set of linked HTML pages with no backend and no build step, offline-first, originally built for an iPad during heavy travel.
Audience
Built first as a working model of a connectivist personal learning environment, for the professional development of the educator who made it. It also serves educators tracking AI in education, since the architecture is deliberately adaptable: change the key variables and the same tool can be rebuilt for a different person, institution, or goal.
Process
The most involved piece is a curated RSS news feed run through a custom keyword classifier that routes each article to the right tab, a multi-proxy fetch waterfall so feeds still load when one proxy fails, and local caching so articles stay readable offline. This was the first AI build here and the experiment that led to everything after it. It is still in use, and the rest of this work grew from what it taught.
Technology
- Six linked HTML pages on one design system: no backend, no login, no build step.
- localStorage for all notes, checkboxes, and the activity log, on the user’s own device.
- A custom keyword classifier over sixty-two RSS sources, with no model at runtime.
- A multi-proxy fetch waterfall with 24-hour local caching, offline-first.
- Hosted on GitHub Pages.
Outcomes
Cultivate proved the pattern the rest of this portfolio is built on. It is the first AI build here, and everything after it, Render, CopaMigo, the Syllabus Checker, grew from what it taught: a single-file, no-backend, offline-first personal learning environment can hold real professional practice without a platform behind it. Success here is measured by continued use rather than a launch metric. It stays in daily rotation as the working reference model of connectivism in practice, and its architecture is deliberately portable, so changing the key variables rebuilds the same environment for a different person, institution, or goal.
Status
Stable and in ongoing use. Version 1.3, the working reference model of a connectivist personal learning environment and the pattern the later tools grew from. It keeps no student data: notes, checkboxes, and the activity log live only on the user’s own device.