Hi, I'm Anubroto

Software Engineer | AI/ML Enthusiast | Web Developer

Profile Picture

About Me

I see technology as more than just a tool—it's a force for change. With a passion for innovation and a curiosity for solving complex problems, I am constantly exploring new ideas and pushing boundaries.

Skills

Technical expertise across various domains.

Python, JavaScript, Rust Flask (Backend) HTML, CSS, Streamlit SQLite, Supabase ML (Scikit-Learn, Pandas) Deep Learning (PyTorch) Data Analytics API Integration Deployment (Vercel) Security & Auth Web Dev (CRUD, MVC)

Projects

IPL Stats Explorer

  • Developed a dynamic web platform to access IPL stats (2008-2022).
  • Implemented customizable data retrieval with Flask backend.
  • Deployed using Vercel for seamless accessibility.
  • Designed efficient search and stats filtering algorithms.
  • Conducted statistical analysis for cricket analytics.

Sales Predictive Analytics

  • Developed a predictive analytics system using Python & Scikit-Learn.
  • Implemented supervised learning techniques to forecast sales outcomes.
  • Utilized historical data analysis for accurate predictions.

Argentic ChatBot

  • Developed an AI chatbot with Streamlit and persistent chat history.
  • Implemented LangChain for conversation management.
  • Integrated OpenAI/Gemini API for contextual responses.
  • Used secure API key management and UUID-based tracking.

Rake

  • Developed a blazing-fast, beginner-friendly HTTP server in pure Rust using only the standard library.
  • Implemented custom route handling and multi-threaded TCP request processing without external crates.
  • Enabled static file serving with MIME type detection and dynamic content support.
  • Designed reusable and extensible server components for easy integration into Rust projects.

VisionForge

  • Built a web-based interface to generate and export custom PyTorch models dynamically using React and TailwindCSS.
  • Implemented a FastAPI backend to convert JSON-based model architecture into PyTorch code and serialize models or weights as downloadable `.pth` files.
  • Enabled separate download options for full model and model weights using the FileSaver.js library for smooth frontend-to-backend integration.
  • Used Pydantic for structured data validation.

Ghosenet

  • Created a tensor computation framework in Rust with automatic differentiation.
  • Implemented support for operations like transpose, abs, conditional select, and broadcasting.
  • Designed a custom computational graph to support backpropagation and gradient accumulation.
  • Added serialization support for saving and loading tensors using JSON files.

Resume

View and download my resume below.

Download Resume

Contact

Let's connect! Reach out via email, LinkedIn, or GitHub.

Contact Me

Email: anubrotoghose@gmail.com

LinkedIn: LinkedIn Profile

GitHub: GitHub Profile