Paul Immanuel

Mathematics Masters graduate from the University of Bonn, Germany. I have a strong base in programming, data science, machine learning and data analytics. Looking for fresher or internship opportunities in the data science domain. I am now based in Ernakulam (Kerala) but am open to relocating if needed.

Experience

Teaching Assistant - Topological Data Analysis

University of Bonn

Bonn, Germany

June 2024 – Sept 2024

• Helped students with foundational mathematical concepts such as homology and co-homology, while running labs that focused on practical applications to machine learning.
• Particular focus on kernel methods that help extract topological information from data sets that is robust to small perturbations using persistence diagrams.

Multi-Timeframe EMA trading stradegy (GitHub) Dec 2025

• Developed a multi-timeframe cryptocurrency trading strategy based on Exponential Moving Averages (EMAs), with consistent signal logic across live Binance execution and historical backtesting.
• Integrated high-frequency (1-second) and low-frequency (1-minute) EMA-based signals to align short-term momentum with higher-timeframe trends.
• Implemented both live trading logic (Binance API) and historical backtesting using backtesting.py to validate execution consistency.

• Created a program that identified instances of backhand shots in videos of matches and training.
• Used MediaPipe pose estimation to extract joint and limb data from videos and used VGG image annotation software for temporal labelling of shots.
• This method drastically reduced the amount of raw data required and greatly simplified the complexity of the model.

Thesis prepared and written over a year long process, supervised by Prof. Jessica Fintzen. Published on the ArXiv.

• Treats the topic of Banach-Colmez spaces, for which we give a detailed introduction to the theory of perfectoid spaces as laid out by Fields medal winning Prof. Peter Scholze in recent work.

Master of Science in Mathematics, University of Bonn, Germany