index cce1211..bf353b9 100644
@@ -2,6 +2,7 @@
status: raw
tags:
- ai
+- education
title: project suggestor
type: idea
updated: 2026-04-11
@@ -10,4 +11,8 @@ visibility: public
# project suggestor
-AI project recommendation based on who you are.
\ No newline at end of file
+an AI that recommends what to build based on who you are — your skills, your gaps, your goals, your interests, and your timeline. the observation is that one of the biggest sources of wasted effort for developers and learners is choosing the wrong project: too easy (nothing learned), too hard (abandoned after a week), wrong domain (doesn't move your goals forward), or too generic (the 100th todo app). a good mentor gives you project suggestions that fit your specific situation; this is that, automated.
+
+the input model: a lightweight profile capturing what you already know, what you want to learn, how much time you have, and what kinds of problems interest you. the AI matches this against a space of project templates and adapts them — it might suggest building a CLI tool in Rust if you're a Python dev who wants systems programming experience, or a data pipeline project if you want ML engineering practice. the suggestions should include reasoning (why this project for you, what you'll learn from it, potential blockers to watch for) not just a project name.
+
+this connects naturally to [[project-designer|project designer]] which takes over once you've chosen a project and builds out the full spec. [[learning-suite|learning suite]] is the broader platform context. [[axon|axon]] could power the personal profile side — if a system already knows your history and skills, project suggestion becomes a much richer recommendation problem. [[intelligence-development|intelligence development]] would deploy both this and project-designer as core products. [[me-model|me model]] is the most ambitious version of the personal profile: a model trained on everything about you would give incredibly precise project suggestions.
\ No newline at end of file