Lars Brestrich
Data Scientist with MSc from St Andrews (Distinction) and 3+ years of professional software development experience. Expertise in statistical analysis, data manipulation and data visualisation, complemented by strong programming skills in Python and C++.
Data Scientist
Technical Skills
Proficient in: Python, C#, C++
Experience in: SQL, R, Java, javascript
Tech Stack: Jupyter, Tableau, RStudio, Excel, Git
Education
M.Sc. Data-Intensive Analysis | University of St Andrews (September 2025) - Grade: Distinction (83%)
• Applied machine learning and statistical techniques including neural networks, random forests, GLMs and GAMs for classification and regression problems.
• Produced 9 end-to-end data science projects that included data manipulation, feature engineering, model development and model evaluation. Implemented solutions using Python (Scikit-Learn, Pandas, NumPy), R, and SQL.
• Conducted dissertation research investigating missing data mechanisms in causal inference using a simulation study with 24,000 variable combinations to inform methodological best practices.
B.Sc. Creative Computing | Goldsmiths University (July 2020) - Grade: First Class Honours (75%)
• Developed code projects in C++, Python, and Java with focus on machine learning and computer graphics.
• Self-taught advanced algorithms, game engine architecture, and real-time systems development.
Work Experience
Programmer @ Bohemia Interactive (Mar 2021 - Mar 2024)
• Developed high-performance algorithms in C++ for real-time systems serving 20+ million users, focusing on user-facing features, performance analyses and optimisation in performance-critical environments.
• Collaborated with cross-functional teams of 20+ programmers, designers, and testers to translate product requirements into technical implementations.
• Followed a high technical standard to ensure readable and maintainable code in a 500K+ line codebase.
Co-Founder @ X-Tron (September 2018 - Jan 2021)
• Built website and collaborated on marketing strategy for the first digital marketplace for export finance, which attracted 100+ corporates and major banks.
Selected Projects (Data Science)
Visualisations of Nobel Laureate Data
A group of D3js data visualisations in javascript answering the question: “Where should I be to win a Nobel Prize?”, aimed at globally curious and ambitious researchers and students. It combines the data of nobel laureate winners from the past 50 years with various country-specific metrics such as GDP, population and political freedom. Group project.
Machine Learning Prediction: Spatial/Temporal Temperature
Trained and optimised Machine Learning models to predict temperature from a ERA5 dataset, which includes spatial and temporal weather data. Modelled KNN, neural network, MLPRegressor, Random Forest and Linear Regression. Group Project.
Exploratory Analysis of US Fortune 500 Companies
Exploratory data analysis of data gathered by Fortune Magazine on key company metrics. Our goal is to uncover relationships between these variables by fitting a linear model, and explore how these findings could be used to inform investment decisions in the future.
Python Game (Fences)
Python implementation of a networked game called “Fences”. Incorporates networking code for client-server architecture, hierarchical user interfaces and game logic.
Competitor Analysis Report
Statistical report comparing the effectiveness of hair products for male-pattern baldness from 4 different competitors. Investigates the comparative effect of products and the age of subjects using t-tests, Tukey’s HSD, and linear modelling.