Jan Krejčí

Building a modular framework automating data pipelines, modeling (Hyperopt), SHAP interpretation, and clustering - adopted across 5+ projects, cutting setup time ~30%. Created a unified campaign monitoring tool, reducing troubleshooting time 40%. Developed partner profiling and churn prediction models with MLflow pipelines and A/B testing. Led data strategy for a €1M Apple project and built onboarding programs for 10+ team members.

Developed and maintained data pipelines and pricing models for Private Property insurance, combining actuarial rigor with automation to improve efficiency and insight generation. Key achievements: Built and maintained GLM-based risk models for Private Property, including technical pricing at the individual policy level. Designed and automated the renewal pricing data process, generating a financial impact in the low millions of CZK. Created an algorithm for estimating sum insured values for flats, applied

Provided analytical and reporting support for actuarial and financial teams, automating key workflows and improving data accuracy in core insurance processes. Key achievements: Automated report generation using VBA macros, significantly reducing manual effort and improving reporting consistency. Designed and implemented a data validation framework for critical management reports, ensuring reliability of key metrics. Developed an Excel-based tool to consolidate non-standardized data sources for annuity repo