How it started
My cousin and my sister both wanted to learn Python. Both are smart, motivated people with no programming background. Both tried the usual routes — YouTube tutorials, free courses — and both stalled.
The pattern was familiar: passive video watching, exercises that don't adapt to where you are, and the moment you hit something confusing, no one to ask. The feedback loop breaks, momentum dies, and the course gets abandoned.
I thought I could build something better in a weekend. It took longer. But the core idea was simple: what if learning Python felt like having a patient tutor next to you while you wrote code?
What I built
CodeCraft Academy is a web app with two main components working together:
In-browser Python editor
No installation, no environment setup, no "first let me configure your PATH variable" friction. Open the browser, write Python, run it, see the output. The editor runs Python directly in the browser, so the barrier to first code is zero.
Conversational AI tutor
While you write code, you can ask the AI anything:
- "What does this error mean?"
- "Explain what a list comprehension is, with an example"
- "Can you give me a small exercise to practice this?"
- "Why does my loop run one extra time?"
The AI responds inline — not by redirecting you to documentation, but by generating a focused explanation or a short lesson tailored to exactly what you asked. It can also generate new exercises on-demand based on where you are in the session.
The key design choice: the curriculum emerges from the conversation, not from a fixed syllabus. You learn what you need, when you need it.
What makes it different
Most AI coding tutors are wrappers around a chat interface with code formatting. The distinction here is the tight integration between the editor and the tutor — when something goes wrong in your code, you are one message away from an explanation, and the explanation references what you just wrote. The loop between writing, running, getting stuck, and understanding closes much faster.
Status
Experimental. Built for personal use — my cousin and sister both used it during the initial version. The core learning loop works. It is not publicly deployed, and the UI is a first pass.
This is a tinkering project, but the underlying idea — conversational AI + live code environment as a learning primitive — feels worth developing further.