Anton Poletaev

• Engaged with senior government stakeholders to identify productivity bottlenecks and potential AI solutions, rapidly prototyping generative AI tools.

• Conducted CoT faithfulness research within a funded training program mentored by Neel Nanda (Google DeepMind)

SigTech is a leading provider of quant technology for macro strategies and cross asset allocation, offering a platform for data ingestion, cleaning, validation, research, and strategy deployment to clients in over a dozen countries, with a combined AUM of over $5 trillion. • Optimized implementation of trading strategies within SigTech's API codebase, including mean reversion, momentum, pairs trading, and calendar spreads; implemented unit tests and data validation frameworks to ensure the robustness and reliability of the platform.

• Developed a novel causality-based semi-supervised feature selection algorithm for the firm’s flagship AutoML platform, utilizing Bayesian inference models via statsmodels and other algorithms through scikit-learn, TensorFlow, and XGBoost. • Executed thorough controlled-environment A/B testing to compare the developed algorithm’s performance to the options currently available on the platform; the former has been shown to outperform available techniques in terms of the final predictions’ accuracy across numerous machine learning methods by an average of 11%. • Authored technical reports and whitepapers detailing the algorithm's methodology and performance metrics.

• Core team member on brokerage app strategy project; managed end-to-end interview workstream spanning 20+ user interviews to identify key behavior patterns and friction points; supported development of market segmentation strategy through qualitative research and conjoint analysis, identifying 4 key user personas and their distinct feature preferences. • Contributed to GTM roadmap prioritizing high-value segments, including channel strategy and targeted value props for each persona. • Designed tiered pricing model aligned to segment willingness-to-pay, projecting 25% revenue growth in first year. • Built framework for educational content platform reaching 10k monthly users, driving improvement in trader retention. • Thorough product and competitor studies, event preparation, slide and video editing.

• Research and evaluation of online banking solutions for retail customers. • Assisting customers with service and product enquiries. • Participated in negotiations for mortgages.

Modules included: - C++ for Computational Finance - Big Data and Applied Methods for Finance (I & II) - Quantitative Investing - Blockchain and Applications - Derivatives Applied Project: “Optimizing Trading Strategies: A Multi-LLM Classification Analysis of Financial News”

- Thesis Title: “Suggesting a Novel Causality-Based Feature Selection Technique for Multivariate Time-Series Data”

- KCL Case Competition 2021 Winner - Thesis Title: “National Macroeconomic Impulse Responses to Oil Price Shocks: A Study Employing Inventories-Refined SVAR Model” - Chairman at the KCL Russian Speaking Society