Priyadip Sau
I am a highly motivated M.Tech student at IIT Jodhpur specializing in Artificial Intelligence. With a strong background in Mathematics, I am passionate about leveraging my analytical skills to develop robust machine learning models and explore the frontiers of generative AI.
View My CV
About Me
My journey into AI wasn't a straight line; it's a calculated collision of abstract mathematics and hands-on $M.Tech$ innovation. I'm currently deep in the trenches, translating the theoretical elegance of my math degrees into tangible, cutting-edge AI solutions.My core belief? The most robust AI isn't just about massive datasets; it's about elegantly structured models. My deep dive into the fundamental principles of computation isn't just academic—it's my toolkit for building AI that is efficient, scalable, and doesn't just work, but works smart.Fueled by a dangerous amount of coffee and the belief that the right equation can solve (almost) anything, my goal is to contribute to AI that is intuitive, impactful, and maybe even a little bit surprising. Let's build something that matters.
Academic Foundation
Education
Indian Institute of Technology Jodhpur
M.Tech in Artificial Intelligence (2025 - 2027)
Relevant Coursework:
University of North Bengal
M.Sc. in Mathematics (2021 - 2023)
Relevant Coursework:
Vidyasagar University
B.Sc. in Mathematics (2018 - 2021)
Relevant Coursework:
My Tech Stack
Python
PyTorch
PEFT
LangChain
XGBoost
LightGBM
Scikit-Learn
Pandas
NumPy
SciPy
Matplotlib
Seaborn
Plotly
Git
GitHub
MLflow
LIamaIndex
C++
VS Code
PyCharm
Jupyter
LaTeX
Experience
Math Teacher
Vivekananda Public School, Odisha
March 2023 – July 2025
Taught mathematics to students of Class-VIII to Class-XII at a CBSE affiliated school.
Manage Network Expert
Chegg India, New Delhi, Delhi 110025
Feb 2020 – Jan 2025
As a Subject Matter Expert, I translated complex problems in Advanced Math and Calculus into clear, step-by-step solutions for a global student base. My role was to deliver academic clarity on-demand, turning textbook panic into practical understanding.
Projects
Rice Dataset Classification using ResNet-18
Trained a ResNet-18 classifier from scratch, achieving 99.87% accuracy on a 5-class rice species dataset. Utilized Python and PyTorch for implementation.
View on GitHubPersona Modeling with Fine-Tuned LLMs
For my Foundation Models & Gen AI course, I fine-tuned the Hermes-2-Pro-Llama-3-8B model to specialize in persona modeling, enhancing its ability to generate contextually consistent character responses.
View on GitHubAdvanced Customer Segmentation
Developed a novel clustering system for an ML course project by combining autoencoders for non-linear feature extraction with Gaussian Mixture Models (GMM) for probabilistic clustering.
View on GitHubGet in Touch
I'd like to hear from you!
I believe the most exciting breakthroughs in AI stem from collaboration. If you are working on challenging problems in machine learning, deep learning, or LLMs, I'd love to connect.