RStudy: Turning University Study Chaos into Active Recall with Local-First AI
AI multiplied information access but increased study overload. How RStudy was built as a local-first desktop app for active recall without recurring subscriptions.
The rapid rise of AI tools has drastically multiplied the amount of information available, yet when exam week approaches, students face familiar friction: overwhelming chaos. Fragmented notes, scattered presentation slides, and passive re-reading routines consume hours without building long-term memory retention.
To solve this problem, I started building RStudy, a local-first desktop application designed to bring order to study sessions by transforming regular notes into active recall systems while keeping personal materials private.
The AI overload paradox and passive learning limits
Instant access to summaries is often mistaken for actual learning. Having a chatbot explain a mathematical theorem does not mean you can solve it under exam conditions.
Cognitive science highlights two critical methods for durable learning:
- Active Recall: challenging the brain to retrieve answers before looking at the solution.
- Spaced Repetition (SRS): scheduling reviews at calculated intervals based on self-reported difficulty.
Most existing SaaS tools introduce two major drawbacks: mandatory monthly subscriptions and forced cloud uploads of proprietary course materials.
Inside RStudy: From raw notes to active quizzes
RStudy unifies the study preparation workflow into a clean, distraction-free environment:
1. Course & Exam Dashboard
A central overview to track academic credits (CFU), exam deadlines, lesson progress, and pending daily flashcard reviews.
2. Note Editor & AI Study Pack
A distraction-free rich-text editor supporting voice dictation. From the AI Study tab, notes are processed into:
- Executive Summaries & Key Points: structured outlines of fundamental principles.
- Mermaid Diagrams: interactive flowcharts and concept maps rendered on the fly.
- Presentation Slides: full-screen decks tailored for oral rehearsal.
- Progressive Exercises: multiple-choice questions, open-ended tests, and code exercises with immediate feedback.
3. Spaced Flashcard Queue
Generated flashcards are automatically organized by subject and difficulty (Easy, Medium, Hard) for efficient daily retention.
Local-first architecture and upcoming launch
RStudy is designed around operational sustainability, privacy, and user ownership:
- Local-first SQLite DB: notes, subjects, and study history stay strictly on your local disk.
- Flexible AI Execution: runs 100% offline via local compact models (llama.cpp) or connects directly to personal OpenAI-compatible API endpoints (e.g. DeepSeek).
- No Recurring Subscriptions: one-time license model.
Development is in its final polishing stages, with the official launch arriving soon.