Sudip Das

AI/ML Engineer

AI/ML Intern

M.Tech Data Analytics @ IIT (ISM) Dhanbad

Specializing in:

|

Building intelligent AI systems with Deep Learning, Federated Learning, and Generative AI

About Me

I'm an AI/ML Intern and M.Tech Data Analytics student at IIT (ISM) Dhanbad, driven by a passion for building intelligent AI systems and privacy-preserving machine learning frameworks.

My research focuses on Heterogeneous Architecture Split Federated Learning, exploring distributed AI training with adaptive model partitioning and privacy-preserving learning techniques.

I combine strong mathematical foundations with practical software engineering to develop real-world AI solutions in Deep Learning, Computer Vision, and Large Language Models.

M.Tech Data Analytics

Indian Institute of Technology (ISM) Dhanbad

M.Sc Mathematics

Visva-Bharati University

B.Sc Mathematics

Ramakrishna Mission Vivekananda Centenary College

Experience

AI/ML Intern

IIT (ISM) Dhanbad

  • Developing AI/ML applications and models for real-world use cases
  • Building and optimizing deep learning models for production deployment
  • Designing backend APIs for AI system integration
  • Implementing distributed AI training frameworks
  • Deploying AI systems on edge devices and cloud infrastructure

AI Research Focus

Privacy-Preserving Distributed Learning

  • Researching heterogeneous architecture split federated learning
  • Developing adaptive model partitioning strategies
  • Implementing privacy-preserving aggregation mechanisms
  • Optimizing distributed training for heterogeneous clients

Research

HA-SplitFed

Heterogeneous Architecture Split Federated Learning

An advanced distributed machine learning framework designed for heterogeneous environments where clients have varying computational capabilities and neural network architectures.

🔐

Privacy-preserving distributed AI

🏗️

Heterogeneous client architectures

🧠

ResNet/MobileNet/CNN support

⚙️

Multi-head server architecture

🎯

UCB-based dynamic split layer selection

📊

Async FedProx aggregation

Projects

HA-SplitFed

Advanced Split Federated Learning framework for heterogeneous AI environments with privacy-preserving distributed training.

Federated Learning Privacy-Preserving Heterogeneous
Python PyTorch Deep Learning
View on GitHub →

ChestVisionAI

AI-powered chest X-ray analysis assistant helping radiologists generate faster medical reports with disease predictions.

Disease Prediction Heatmap Visualization Local Deployment
Deep Learning Computer Vision FastAPI
View on GitHub →

Local LLM Chat System

Private AI chatbot system using locally deployed LLM models with secure authentication and FastAPI backend.

Ollama Integration Secure API Authentication
Ollama FastAPI LLM
View on GitHub →

Skills & Expertise

Programming

Python
C++
SQL

AI/ML

Machine Learning
Deep Learning
Neural Networks
Computer Vision
Generative AI

Frameworks

PyTorch
TensorFlow
Scikit-Learn

AI Tools

Ollama
LLMs
LangChain

Backend

FastAPI
Streamlit

Tools & DevOps

Git
GitHub
Docker
Linux

Achievements

🏆

GATE Mathematics 2024

Qualified

🎓

JAM Mathematics 2021

Qualified

JEE Main 2017

Qualified

Get in Touch

I'm always interested in hearing about new projects and opportunities.

Note: This is a static form. To make it functional, integrate with a backend service.