SynapseLearn AI

📌 1. Educational Problem & Social Impact

Online education suffers from a fundamental pedagogy flaw: passive content consumption (watching 40-minute video lectures or reading dense PDF textbooks) yields an 85% drop-off rate and poor long-term conceptual retention:

• Passive Illusion of Competence: Students believe they understand a concept while reading, but experience complete mental block when asked to apply it or reason from first principles.
• Rigid One-Size-Fits-All Explanations: Traditional tools explain concepts at a single arbitrary difficulty level, failing both beginners seeking intuition and advanced majors requiring mathematical rigor.
• Premature Answer Spoilers: Standard chatbots immediately give direct answers, robbing students of the cognitive struggle required for neuroplastic learning.

We built SynapseLearn AI—an adaptive multi-modal educational platform that transforms how people learn, teach, and internalize complex knowledge through guided Socratic inquiry, interactive mental model parameter testbeds, and multi-tier Feynman simplification.

🌟 2. Key Features

1. 5 Multi-Modal Mental Model Sandboxes:
   • ⚛️ Quantum Mechanics: Interactive Double-Slit experiment. Adjust photon wavelength $\lambda$ and slit distance $d$. Toggle the slit observer detector to witness the live quantum wavefunction collapse into classical particle bands!
   • 🧠 Deep Learning & Neural Networks: Gradient descent testbed. Adjust learning rate $\alpha$ and epochs. Watch the computed loss function update in real-time, with automatic alerts if the learning rate causes gradient divergence ($\text{Loss} \rightarrow \infty$).
   • 🧬 Genetics & Molecular Biology: CRISPR-Cas9 sequence matcher. Adjust guide RNA complementarity and trigger targeted double-strand breaks with HDR (Homology-Directed Repair) vs NHEJ pathways.
   • 🪐 Astrophysics: Black hole mass slider ($M_{\odot}$). Dynamically calculates the Schwarzschild event horizon radius ($r_s = \frac{2GM}{c^2}$) and gravitational escape velocity.
   • ⚡ Distributed Systems: 5-Node Raft Consensus cluster. Click any node to crash or reboot it; dynamically computes quorum ($\lfloor N/2 \rfloor + 1$) and cluster safety.
2. 🎙️ Web Speech Socratic AI Voice Tutor: Native browser bi-directional voice dialogue powered by the Web Speech API (SpeechRecognition & SpeechSynthesis). Detects conceptual misconceptions and responds with guided scaffolding questions rather than giving away answers directly.
3. 🧒 4-Tier Feynman Technique Simplifier: 1-Click dynamic shift between ELI5 (5-Year-Old), High School AP, Undergraduate, and PhD Researcher levels with real-world analogies and key takeaways.
4. 🏆 Active Socratic Recall & Spaced Repetition Mastery: Interactive concept checks with multiple-choice diagnostic options, instant feedback on reasoning gaps, and celebration confetti on mastery gain (+15 XP).
5. 📑 1-Click Study Dossier & Anki Flashcard Exporter: Export complete study guides to Markdown (.MD), Anki-compatible JSON flashcards, and Printable A4 PDF Cheat Sheets.

🛠️ 3. Technologies Used

• Frontend & UI: React 19, TypeScript, Vite 6, Tailwind CSS 3.4
• Voice Speech Pipeline: Web Speech API (SpeechRecognition & SpeechSynthesis)
• Interactive UX: Canvas-Confetti + Lucide Icons
• License: MIT Open-Source License

Built With

• artificial-intelligence
• github
• machine-learning
• react
• tailwindcss
• typescript
• vercel
• vite
• web-speech-api