Shuban M S

Hello, I am Shuban.

I am a
My Resume

About Me

I am a final-year Computer Engineering student at VIT Chennai with a deep passion for AI, machine learning, and data science. With three years of hands-on experience in data analytics and deep learning projects, I've developed a strong foundation in Python development and published research in IEEE and Springer journals. My expertise spans across Python, MySQL, PostgreSQL, and Azure cloud services, with a particular focus on Deep Learning and Generative AI applications.

I have a strong research background and am quick to learn and implement new technologies, which allows me to adapt rapidly to emerging trends in the field.

As a Data Scientist and AI Engineer, I've demonstrated my capabilities by reaching the finals in three National Hackathons and one International Hackathon. I thrive on solving complex problems and am constantly exploring new developments in AI technology. As I approach graduation, I'm eager to take on new challenges and contribute to innovative projects in the field of artificial intelligence.

My ability to quickly grasp new concepts and apply them effectively has been key to my success in both research and practical projects.

My Skills

Tech Stack

PostgreSQL
Redis
Pandas
NumPy
Scikit-learn
TensorFlow
PyTorch
Jupyter

Technologies and Frameworks

FastAPI
Docker
Azure
VS Code
LangChain

Projects

Achievements

Hackathon Finalist

Finalist in 3 National Hackathons and 1 International Hackathon.

Patent

Submitted a patent on a Generative AI learning system with Blockchain integration.

Speaker

Conducted a session on Azure Fundamentals and Deploying Machine Learning Models on Azure.

Research Papers

FALCONNet: A Multi-Defense Approach for Securing Few-Shot Learning Against Adversarial Attacks

This paper compares few-shot learning models—Siamese, Prototypical Networks, and our proposed FalconNet—on the Omniglot dataset. We assess their robustness against adversarial attacks (PGD, FGSM) and test defense strategies like adversarial training and defensive distillation. FalconNet, with these defenses, outperforms the other models in accuracy and stability under attack.

Contact

Email: msshuban4@gmail.com

LinkedIn: linkedin.com/shubanms

GitHub: github.com/shubanms