Kirill Safonov is a researcher and data scientist with experience in applied machine learning and quantitative analysis. He has worked as a Research Assistant at Columbia Business School and Harvard Business School, and as a Data Scientist at Yandex. His current work focuses on AI applications in quantitative finance and price prediction in competitive, uncertain markets.

I’m interested in AI agents, especially their use in finance for trading, decision-making, and optimization. I want to build and play with them to see what actually works. Curious to connect with others exploring agent-based systems or applying RL in real-world settings.

1. AI Agents in Quantitative Finance: Designing and evaluating AI agents for adaptive trading and portfolio optimization using market data. 2. Price Prediction in E-Grocery: Building ML models to forecast product prices in e-grocery markets, accounting for competitor pricing and demand volatility.

Kirill Skobelev is a research professional with a strong background in economics and AI. He graduated summa cum laude from the New Economic School and Higher School of Economics. Kirill has worked as a research assistant at the University of Chicago and Stanford Graduate School of Business, and as a data scientist at Plata Card. He is proficient in multiple programming languages, including Python and R, and has received several awards for his academic achievements.

Finance, scientific discovery with AI, LLMs+tools.

- Trained a StyleGAN model using >2500000 scraped food images and applied a method to backout survey partipants' perceptions of food. - Extracted commodity price expectations from over 400k investor call transcripts, achieving 90% precision against human labels. Worked on token-efficient methods for compressing few-shot examples as a subproject. - Built and integrated ML forecasts into a NPV model at Plata Card (Mexican fintech unicorn) to improve portfolio decisions.