CS student at Islington College, Kathmandu. Got into programming out of curiosity and now I can't stop — I spend most of my time building stuff with Python, ML, and IoT. Occasional fights with dependency hell, and I celebrate when things compile on the first try (rare, but it happens).
- 🔭 Just finished ParkX — an IoT parking system. Wiring an ESP32 at 2am is an experience.
- 🌱 Going deeper into Deep Learning (PyTorch) and RAG apps (LangChain) — slowly, but getting there.
- 💡 I learn by doing. Lost count of how many projects I started just to learn one thing.
- 📫 Reach me at Praansu12@gmail.com — always happy to chat.
|
IoT parking system with ESP32 + FastAPI. Real sensors, real dashboard, real late nights debugging. |
PyTorch CNN with transfer learning. 4K+ images, way too many hyperparameter tweaks. |
|
Upload PDFs, ask questions. LangChain + ChromaDB + OpenAI under the hood. Still amazed this works. |
XGBoost + SHAP for heart disease prediction. Because accuracy means nothing if you can't explain why. |
|
PyQt5 app for classifying vehicle images. Keyboard-first — built because manual labeling drives me crazy. |
E-commerce frontend for eco products. Pure HTML/CSS/JS — no frameworks, no build tools. |
| Project | Link |
|---|---|
| 🌿 EcoVerda | praansu.github.io/eco-verda |
| 🧮 JS Calculator | praansu.github.io/js-calculator |
| 👤 Portfolio | praansu.github.io |
Languages — Python, Java, JavaScript, HTML/CSS, SQL, C++
ML/AI — PyTorch, scikit-learn, Pandas, NumPy, LangChain, OpenCV, XGBoost
Backend & Tools — FastAPI, Streamlit, Flask, Git, PyQt5, SQLite
IoT & Cloud — ESP32, Arduino, AWS, QGIS
Deep learning with PyTorch (CNNs, transfer learning), RAG-based apps with LangChain, and trying to get better at computer vision. The more I code the more I realize how much I don't know — keeps things interesting.