Tim Golovenko

Computer Science @ Durham University

Building systematic trading models, dashboards and high-performance simualtion engines

About

I'm a Computer Science student at Durham University, building projects centralised around trading, game theory and system optimisation. Currently looking for oportunitites to combine my interest in financial markets with my passion for statictical modeling, (hopefully) helping generate profitable trading strategies.

I'm particularly interested in the intersection of machine learning, statistical modeling, and execution optimization. Through projects such as backtesting tennis betting models and vectorizing Monte Carlo simulations, I enjoy tackling the challenge of building systems that are both accurate and fast.

Tim Golovenko

Experience

Jul 2025 - Aug 2025

Data Science Intern

BT Group (Active Intelligence)
London, UK
  • Validated HGV detection models against DfT freight data across 40+ UK regions, improving accuracy by 10%
  • Performed regression-based error analysis on 10,000+ geospatial observations to identify bias and failure modes
  • Translated probabilistic outputs into deployment decisions for logistics and transport planning tools
Jul 2022 - Aug 2024

Co-Founder & Backend Developer

Carbon Foodprint
Remote
  • Co-founded a sustainable dining rewards platform, designing a data-driven rewards system under cost and adoption constraints
  • Built backend systems in C# and SQL, including QR receipt tracking and spend-to-reward pricing logic
  • Onboarded 5+ beta sites and analyzed incentive sensitivity and misuse edge cases

Education

MEng in Computer Science

Durham University
First Class / GPA 3.8 Expected: 2028
Durham, UK
Probability Statistics Optimisation Algorithms Numerical Methods

Projects

Tennis Betting Model

2025

An end-to-end systematic trading system for tennis match betting. The pipeline ingests historical match data, trains XGBoost models to predict match outcomes with confidence scores, converts probabilities to fair odds, and executes bets when market prices deviate from model valuations.

94% Simulated Return
37 Bets Placed
Python pandas XGBoost scikit-learn

Blackjack/Poker EV Engine

2025

A high-performance Monte Carlo simulation engine for card game analysis. Models complete game rules, shoe penetration, and player policies to estimate expected value, variance, and confidence intervals for different strategies.

3.2M Hands/Min
3.8× Speedup
Python Monte Carlo MDP NumPy

Contact

Email
timgolovenko@icloud.com
Phone
+44 7876 488311
GitHub
github.com/1timp1
Location
Basingstoke/London/Durham, UK
Tim Golovenko