Summary
Ivan Eroshkin is a data science leader with roughly a decade of experience translating financial modelling expertise into production-ready ML solutions, currently heading the Validation unit at Sberbank. He combines fast, pragmatic coding and low time-to-market with rigorous estimation of models' PnL/NPV impact, routinely delivering alternative models that outperform incumbents in ~25% of cases. His background spans tabular modelling, AB/AAB testing, uplift and campaign scoring, as well as CV, NLP (incl. HF pipelines), GANs and MLOps, with strong hands-on Spark and containerized microservice experience. With ~11 years in financial modelling and management experience across teams of 3–15, he pairs domain depth in banking with practical validation methodology authorship. Colleagues describe him as creatively critical and team-focused, valuing robust teamwork as a multiplier for technical value. Based in Russia, he brings a rare blend of finance-first intuition and applied ML engineering that prioritizes measurable business outcomes.
9 years of coding experience
4 years of employment as a software developer
CFA Institute
Lomonosov Moscow State University