Stepan Kropachev

Spearheaded AI strategy, building an R&D unit and managing 13 engineers. Deployed a secure local LLM and "M2 Typewriter," cutting content time by 43%. Architected a RAG agent automating 30% of inquiries. Previously led Core Search, boosting conversions 27% via technical SEO, while fostering a product-led culture and engineering ownership.

Architected product vision and technical strategy for floor-plan recognition AI, built an ML team and delivered an MVP, and led go-to-market pilots with PropTech partners to validate commercial potential.

Transformed an MVP into an enterprise SaaS platform, tripling revenue and scaling product x15 through architecture decisions and 29 banking partnerships that opened new API-driven revenue streams.

Led product vision and execution for a digital ecosystem for 'Green Card' insurance processing, culminating in a strategic acquisition by a major insurance company.

Grade: 90/100

Activities and societies: • Researched the limitsations of LLMs' reasoning abilities in Tax Legal domain • Investigated user experience challenges and model performance in Retrieval-Augmented Generation (RAG) AI products through technical and user-centric research. • Developed and implemented evaluation pipelines to assess model performance against a dual set of technical metrics and end-user experience criteria.

Master's in Human-Centered AI completed in 2025 focusing on human-centered approaches to artificial intelligence.

Grade: Compleated

Activities and societies: • Conducted a multi-year comparative analysis of the legislative frameworks and policy outcomes of the immigration systems of the United States, Canada, and Australia. • Employed a qualitative legal and policy analysis framework, examining primary legislative documents, governmental reports, and integration outcome data.

Admitted to the PhD program in Political Science beginning in 2016.

Specialist Degree in Political Science completed in 2013.