Every session follows the same proven arc. Three phases, each building on the last.
The AI tutor introduces concepts through Socratic dialogue. No lectures. No walls of text. A conversation that meets you where you are.
Ask questions freely. The tutor responds with questions of its own, guiding you toward understanding.
Sandboxed exercises matched to your level. The AI tutor watches, guides, but never gives answers. Difficulty adapts in real time based on your rolling performance window.
"Think about how you might break this into smaller steps. What's the first thing you need?"
"You'll want to iterate through the items here. What type of loop works best when you need the index?"
"result = []\nfor i, item in enumerate(items):\n # What should happen to each item?"
Lesson mode orchestrates all three phases into a guided learning path. Introduction, exercises, then flashcard reinforcement. One flow, three modes working together.
SM-2 spaced repetition schedules reviews at optimal intervals. Each card tracks its own ease factor. Easy cards fade into long intervals. Hard cards come back tomorrow.
Overdue cards surface first. Then today's dues. Then new material, up to your daily limit.