Hardware
💻
Primary Laptop
Windows 11 machine with Nvidia GTX 1650 4GB GPU — handles local model training and inference with CUDA optimisation tricks.
Personal🎮
Nvidia GTX 1650 4GB
Local GPU for fast iteration. Runs Stable Diffusion with float16 + attention slicing, LSTM training, and computer vision model testing.
CUDA 11.x🖱️
Dual Monitor Setup
Code on the left, output/docs on the right. Essential for debugging model training runs while reading papers simultaneously.
PersonalDev Tools
🔵
VS Code
Primary editor. Extensions: Python, Pylance, Jupyter, GitLens, GitHub Copilot, Material Icon Theme.
Free📓
Jupyter Notebook
For exploratory data analysis, model prototyping, and creating reproducible research notebooks.
Open Source🐍
Miniconda / Conda
Environment management. Each project gets its own conda env — keeps dependencies clean and reproducible.
Open Source🐙
Git + GitHub
Version control for all projects. GitHub Actions for CI/CD on portfolio. GitHub Pages for static demos.
Free🟠
PyCharm (secondary)
Used for larger Flask/Django projects where the full IDE features (debugger, database tools) shine.
CommunityML Stack
🔶
TensorFlow / Keras
Primary deep learning framework. Used for all LSTM, CNN, and transfer learning projects. Keras API for rapid prototyping.
Open Source🔥
PyTorch
Used for Stable Diffusion (Diffusers library) and HuggingFace Transformers. Preferred for research-style experimental code.
Open Source🤗
HuggingFace
FLAN-T5 for the HR Automation project. Transformers library for NLP tasks and pre-trained model access.
Free tier📊
Scikit-Learn
Baseline models, preprocessing pipelines, metrics, and classical ML. Always the first benchmark before deep learning.
Open Source🌊
Streamlit
Go-to for ML app demos. Built Car Price Prediction, Text-to-Image Generator, and internship deployments with it.
Open Source👁️
OpenCV
Face recognition preprocessing, lane detection (Canny + Hough), and mosquito wing image processing pipelines.
Open SourceCloud & GPU
🏆
Kaggle Notebooks
Free T4/P100 GPU (16GB VRAM) — primary cloud GPU for training large models. 30 hours/week free usage.
Free🌐
Google Colab
Quick experiments and sharing notebooks. Free T4 GPU for smaller training runs and demos.
Free tier☁️
AWS (learning)
Exploring EC2 for deployment, S3 for dataset storage. Currently studying AWS Solutions Architect concepts.
Pay-per-use🚀
Netlify
Hosts this portfolio. Free tier with auto-deploy from GitHub. Zero-config HTTPS and custom domain support.
FreeProductivity
📝
Notion
Research notes, project planning, paper reading summaries, and IEEE submission tracking.
Free tier🔍
Google Scholar
Literature review and citation tracking for IEEE papers. Set up alerts for LSTM weather forecasting and mosquito detection research.
Free📄
Overleaf
LaTeX editor for IEEE paper writing. Clean collaboration with supervisor for ICRITO and ICIPTM submissions.
Free tier💬
Discord / WhatsApp
Research group communication and project team coordination. Discord for developer communities and hackathon teams.
Free