Junior Data Scientist · Python · Machine Learning · Computer Vision
I learn by building — real projects, from stock screeners to medical imaging models.
I'm a Junior Data Scientist building a strong foundation in ML by shipping real projects:
- 🐍 Python-first — Pandas, scikit-learn, TensorFlow/Keras, PyTorch, OpenCV
- 📈 Finance + ML — built a real-time Nifty 100 stock screener with ML signal filtering
- 🩺 Medical AI — currently competing in the RSNA 2026 knee-MRI Kaggle challenge
- 👁️ Computer Vision — face detection, face swap, tampering detection, image classification
- ✨ Generative AI — Gemini-powered apps & Stable Diffusion pipelines
- 🌱 Currently learning — MLOps, model deployment, and LLM engineering
| Project | Live URL | What it does |
|---|---|---|
| 💧 Image Watermark Tool | karansaini260.github.io/watermark | Upload → add text/logo watermark → download (100% in-browser) |
| 🍄 Mario Game | karansaini260.github.io/mario | Playable platformer with physics engine + auto-play AI |
| 🪪 PAN Card Tampering | karansaini260.github.io/PAN-Card-Tampering-Detection | Tamper detection with OpenCV + SSIM |
| 🔄 Face Swap | karansaini260.github.io/Face-Swap-App | Face swap with OpenCV + dlib |
| 🏠 Projects Hub | karansaini260.github.io | All my projects in one place |
| Project | Description | Stack |
|---|---|---|
| 📈 NSE Screening | Real-time Nifty 100 screener — SMMA crossovers + ML signal filtering | Python · Tkinter · scikit-learn |
| 🦵 RSNA Knee Detection | Kaggle: detect 12 knee abnormalities from MRI + radiology reports | PyTorch · pydicom · Multimodal |
| 💬 Gemini AI App | Streamlit chat app powered by Google Gemini | Python · Streamlit · Gemini |
| 😀 Face Detection App | Real-time face detection with OpenCV | Python · OpenCV |
| ✅ Angular Todo List | Feature-rich to-do app | Angular · TypeScript · Bootstrap |
| 📊 Trader Behavior Analysis | Deep EDA of trader data → insights + PDF report | Python · Pandas |
I'm actively looking for Junior Data Scientist / ML Engineer opportunities and
love collaborating on open source. My repos are MIT-licensed with contribution
guides and good first issue labels — pull requests welcome!
If you'd like to co-develop a project or just chat about ML, open an issue or reach out!
⭐ Explore my repositories and star anything you find interesting!