I'm Nandan K S a passionate Full-Stack Developer and MCA student at CMR Institute of Technology, Bengaluru.
I build scalable web applications and AI-powered solutions that solve real-world problems from real-time video conferencing platforms with on-the-fly meeting summaries to fully offline voice assistants that keep your data private.
I'm captivated by the intersection of real-time systems, AI / NLP, and clean software architecture. For me, writing great software isn't just about making things work it's about making them work beautifully.
"Let's build something impactful together."
- 🎓 MCA Student CMR Institute of Technology, Bengaluru
- 🔭 Currently building AI-integrated full-stack applications
- 🌱 Exploring real-time systems, WebRTC, NLP & AI integration
- 💡 Passionate about privacy-first tech, developer tools, and elegant UX
- 🤝 Open to collaborations, new opportunities, and interesting ideas
Languages
Frontend
Backend
Databases
AI & NLP
Tools & DevOps
An intelligent pipeline that transforms any YouTube video into AI-generated summaries, mind maps, and quizzes using Google Gemini + Flask.
| Feature | Detail |
|---|---|
| 📝 AI Summary | Adjustable length (10–50%), cached with SQLite TTL |
| 🗺️ Mind Map | Mermaid-format visual maps generated from transcript |
| 📋 Quiz | Auto-generated Q&A from video content |
| 🔌 Chrome Extension | In-browser video analysis without leaving the tab |
| ⚡ Real-time Updates | Socket.IO processing progress indicators |
Python · Flask · Google Gemini API · yt-dlp · Socket.IO · SQLite
A full-featured video conferencing platform with real-time AI transcription, smart meeting summaries, and ML-based presence detection.
| Feature | Detail |
|---|---|
| 📹 HD Video | WebRTC peer-to-peer video & audio with screen sharing, grid/speaker views |
| 🤖 AI Summaries | Auto-generated meeting summaries & action items via Google Gemini |
| 👤 Face Detection | MediaPipe ML — auto-away detection + missed conversation catch-up |
| 💬 Live Chat | Real-time messaging with typing indicators and emoji support |
| 🗄️ Persistence | MongoDB meeting & chat storage, full REST API + WebSocket events |
React 18 · Node.js · WebRTC · Socket.IO · MongoDB · Google Gemini AI · MediaPipe
Broadcast audio instantly over the web. One device streams; everyone listens through their own headphones — zero Bluetooth, zero extra hardware.
| Feature | Detail |
|---|---|
| 📡 Audio Capture | Browser Tab Audio (getDisplayMedia) or Microphone (getUserMedia) |
| 🔗 Instant Sharing | Unique 7-character room code + QR code generated on demand |
| 🎛️ Host Controls | Pause, resume, stop broadcast, remove individual listeners |
| 🔒 Security | Helmet headers, CORS restriction, rate limiting, stale-room cleanup |
| 🌐 P2P Transport | WebRTC direct audio — no audio data ever touches the server |
React · Vite · TailwindCSS · Node.js · Socket.IO · WebRTC · Web Audio API
A 100% offline, privacy-first Windows voice assistant powered by OpenAI Whisper and a 4-layer NLP pipeline. Your voice, your control, your data.
| Feature | Detail |
|---|---|
| 🎤 Offline STT | OpenAI Whisper or Vosk — no internet, no cloud, no data collection |
| 🧠 4-Layer NLP | Exact → Fuzzy → Phonetic → Semantic matching with weighted scoring |
| 🛡️ Safety | 4-gate validation + command sandboxing (whitelist-only execution) |
| 🖥️ System Control | Open/close any app, smart window & tab control, 40+ keyboard commands |
| ⚡ Performance | Threaded execution, interruptible TTS, startup prewarming, accent support |
Python · OpenAI Whisper · PyAutoGUI · pyttsx3 · NLP Pipelines · Windows OS
⭐ Thanks for stopping by! Feel free to browse my repos, star something useful, or reach out if you'd like to collaborate on something impactful.