Machine Learning Engineer

Job Summary

We are seeking an experienced Machine Learning Engineer to design, build, and deploy production-grade models for demand forecasting, customer churn prediction, and inventory optimization. You'll work with large-scale transactional data (e.g., orders, customer behavior) to create robust systems that predict rental demand, identify at-risk customers, and manage inventory efficiently, including handling returns and refurbishments. This role is ideal for someone passionate about e-commerce/retail analytics and proficient in Python-based ML workflows.Key Responsibilities

• Demand Prediction: Develop and implement time-series forecasting models (e.g., using Prophet, ARIMA, or LSTM) to predict rental demand by product (SKU), category, and city. Incorporate features like seasonality, holidays, promotions, and external factors (e.g., weather, economic indicators) to achieve high accuracy.
• Churn Prediction: Build classification models (e.g., XGBoost, Random Forests) to predict customer churn based on subscription history, order patterns, and behavioral features. Use outputs to inform retention strategies and integrate with inventory models (e.g., estimating returns from churned users).
• Inventory Management: Design optimization models (e.g., using PuLP or linear programming) to manage stock levels, reorder points, and refurbishment cycles, leveraging demand and churn forecasts to minimize stockouts and overstock costs.
• End-to-End ML Pipeline: Create data pipelines (ETL) for ingesting and preprocessing order data (e.g., from CSV sources with timestamps, SKUs, cities). Feature engineering: Generate 50-100+ features like lagged orders, customer tenure, day-of-week effects, and holiday flags.
• Model Deployment & Monitoring: Deploy models as APIs (e.g., using FastAPI, Docker, Kubernetes) for real-time predictions. Implement monitoring for model drift and retraining workflows. Conduct A/B testing and evaluate models using metrics like RMSE (for demand), AUC-ROC (for churn), and cost savings (for inventory).
• Scalability & Experimentation: Optimize models for large datasets (e.g., millions of orders) using cloud platforms (AWS/GCP). Experiment with advanced techniques like reinforcement learning for dynamic pricing tie-ins.
  

Required Qualifications

• Education: Bachelor's or Master's degree in Computer Science, Data Science, Engineering, or a related field.
• Experience: 3-5+ years as an ML Engineer or similar role, with hands-on experience in retail/e-commerce analytics (e.g., demand forecasting, churn, inventory).
• Technical Skills:
• Proficiency in Python (pandas, NumPy, scikit-learn) and ML libraries (Prophet, XGBoost, TensorFlow/PyTorch).
• Time-series forecasting (ARIMA, Prophet) and optimization tools (PuLP, SciPy).
• Data pipelines (Airflow, Spark) and deployment (Docker, Kubernetes, AWS SageMaker).
• SQL for data querying and cloud computing (AWS/GCP/Azure).
• Soft Skills: Strong problem-solving, ability to work in a small team, and experience with Agile/Scrum methodologies.
• Domain Knowledge: Familiarity with subscription/rental models (e.g., handling returns, refurbishments) in e-commerce.
• Preferred Qualifications
• Experience with reinforcement learning or advanced optimization for dynamic pricing.
• Knowledge of big data tools (e.g., Hadoop, Spark) for scaling models.
• Publications or projects in retail predictive analytics
• Familiarity with RentoMojo-like platforms or the Indian e-commerce market.
  

Skills: data tools,python,demand forecasting,inventory optimization,model deployment,sql,data pipelines