Nilanjan Dey

Commercial Data Analyst @Hastings Direct

London, GB

Worked in uplift models using causal inference techniques (propensity score matching, treatment-control segmentation) to evaluate pricing incentive effectiveness reducing inefficient discount spend Applied Optuna for hyperparameter optimization with nested time-series cross-validation, achieving reduction in RMSE across monthly cycles.Delivered explainable AI solutions using SHAP and deployed interactive dashboards (Streamlit + Power BI) that surfaced model insights to pricing, underwriting, and commercial stakeholders.Automated retraining pipelines with version control via MLflow and integrated batch scoring with Snowflake to support real-time pricing analytics.

Purdue University

Postgraduate Program, Data science

Kalinga Institute of Industrial Technology, Bhubaneswar

Bachelor of Technology (BTech), Electrical, Electronics and Communications Engineering

Queen Mary University of London

Postgraduate Degree, Big Data Science