I need a straightforward churn-prediction prototype built from a sample dataset that contains customer demographics, usage patterns and transaction history. After an initial round of data cleaning and exploratory analysis, basic feature engineering should follow—calculating tenure, flagging complaints, creating usage metrics—so that a Logistic Regression model can reliably estimate each customer’s likelihood to leave. Once the model is trained, I’d like the results surfaced in an interactive Power BI dashboard (my preferred tool). If you happen to work faster in Tableau that’s fine too, but Power BI is what I’ll ultimately share with stakeholders, so please be ready to publish there. To wrap everything up, summarise the key churn drivers the model uncovers and translate them into clear, actionable retention recommendations in a concise report. Deliverables • Cleaned dataset and reproducible code (Python or R) • Trained Logistic Regression model file • Power BI dashboard with churn risk visuals and driver insights • Brief report outlining findings and retention recommendations The prototype should be lightweight yet well-documented so the business team can iterate on it later without heavy rework.