Hariprashad Ravikumar

Projects

More projects coming soon — stay tuned!

• AI-DataScience-Lab: Web-Based Forecasting App (GitHub Repository)
  Built a full-stack forecasting tool to explore the complete data science pipeline, from preprocessing to model deployment. The app supports csv uploads, visualizes trends using matplotlib, fits a scikit-learn regression model, and summarizes the dataset using OpenAI’s GPT-3.5 API. The backend (Flask) is hosted on Microsoft Azure App Service, while the frontend is deployed via GitHub Pages. Version 1.0.0 (Beta) currently supports linear regression, with future plans to add polynomial, ridge, and time series models for enhanced forecasting. You can access the frontend of the app at: AI DataScience Lab.
• HariBot: AI Chatbot for Website (GitHub Repository)
  Developed and deployed a custom AI chatbot using the OpenAI API to enhance user interaction on my personal website. The backend, implemented in Python (Flask), is hosted on Render for scalable cloud performance, while the frontend is integrated into my GitHub Pages for a responsive, cross-device experience. The chatbot delivers real-time responses about my background, research, and professional experience.
• Neural Network from Scratch with NumPy (GitHub Repository)
  Built a two‑layer neural network entirely in NumPy, using ReLU activation in the hidden layer and softmax at the output. Trained on 5,000 examples with a learning rate of 0.1, it reached ~80 % accuracy in just 60 iterations. See the repository’s README.md for a detailed walkthrough of forward/backward propagation and weight updates.
• Automated LaTeX CV Build & Deployment (GitHub Repository)
  Built and deployed an automated GitHub Actions workflow to compile my LaTeX CV from Overleaf into a PDF with each push and deploy it to GitHub Pages, directly linking it to my website repo for seamless access. The workflow leverages CI/CD principles and Bash scripting to automate the process, ensuring reproducible and efficient publishing.
• \(\mathbb{Z}_2\) Lattice Gauge Monte Carlo Simulation (GitHub Repository)
  Developed a Python simulation of \(\mathbb{Z}_2\) lattice gauge theory using Markov chain Monte Carlo methods and Metropolis algorithms. Explored confinement phenomena through Wilson loop measurements and benchmarked results against analytical predictions. Served as foundational computational experience in stochastic sampling and lattice QCD simulation ahead of full PhD research.