Skip to main content
Posted 05 June, 2026

Data Scientist - Forecasting & Pricing

KPI Partners
Pune District, MH, IN Full Time
Reference: 7ed3613290326132

Job Description

Key Responsibilities • Demand Forecasting: Design, build, and deploy scalable demand forecasting models (time-series, ML-based) to predict product demand at SKU, category, channel, and regional levels. • Discount & Price Simulation: what-if simulation tools to optimize discount strategies and maximize margin. • End-to-End Model Ownership: Own the full ML lifecycle—data exploration, feature engineering, model training, validation, deployment, monitoring, and iteration.

• Production Deployment on AWS: Build, train, and deploy models using AWS SageMaker; manage pipelines, endpoints, and model versioning in cloud-native environments. • Stakeholder Collaboration: Translate complex analytical outputs into clear, actionable insights for business leaders; present findings and recommendations to senior leadership. • Power BI: Create automated reports to present and track demand forecast model output.

• Data Pipeline Development: Collaborate with Data Engineers to build robust, scalable data pipelines supporting model training and inference. Must-Have Skills • 6–8 years of hands-on experience in Data Science, ML, or Advanced Analytics • Strong experience in Demand Forecasting (ARIMA, Prophet, LSTM, XGBoost, or similar) • Proven expertise in Pricing/Discount Simulation (price elasticity modeling, scenario analysis) • Must have deep understanding of at least couple of Retail/CPG use cases such as customer segmentation, recommendations, demand forecasting, sentiment analysis, inventory optimization, promotion uplift modeling, campaign analysis, churn prediction, etc. • Hands-on production experience with AWS SageMaker (model training, hyperparameter tuning, deployment, batch/real-time inference) • Programming: Advanced Python (pandas, NumPy, scikit-learn, TensorFlow/PyTorch); SQL for data extraction and transformation • Statistical & ML Techniques: Regression, classification, time-series forecasting, ensemble methods, feature engineering

Sign up for Job Alerts