Kamil Ahmadov

Engineered a robust pricing model for CASCO insurance using Generalized Linear Models (GLM) to analyze client data and predict potential losses, enabling more accurate policy pricing. Developed and deployed a propensity-to-buy model on the Dataiku platform with monthly real-time data preprocessing and iterative retraining, doubling the conversion rate and boosting sales of voluntary property policies. Contributed to a computer vision model for detecting auto-damage parts to support claims assessment, and overhauled motor pricing models and APIs, including a redesign of the personal client discount calculator (NCD), reducing response time from 20 seconds to 300 milliseconds and minimizing computational resource usage.

Implemented A/B testing strategies to optimize the user experience of an online platform, resulting in increased player retention. Designed a machine learning model for player churn prediction, achieving a 20% reduction in average quarterly churn and driving improved retention and revenue. Presented findings from exploratory data analysis of game presenters’ behavior, delivering actionable insights to identify areas for performance improvement.

Implemented A/B testing strategies to optimize the user experience of online platform, resulting in an increase in player retention. Designed a machine learning model for player churn prediction, resulting in a 20% reduction in average quarterly churn and driving improved customer retention and revenue growth. Presented findings from exploratory data analysis of game presenters' behavior, delivering actionable insights to identify areas for performance improvement.