Senior Data Scientist - Demand Forecasting

Mai Placement is recruiting for a growing, privately held manufacturing company seeking a hands-on Senior Data Scientist specializing in Python, machine learning, time-series forecasting, and predictive analytics.

The primary objective is to build and improve forecasting models that predict physical-product demand at the SKU level, helping the business make better inventory, purchasing, replenishment, and production decisions.

We are looking for someone who has personally developed forecasting models using Python, understands how to measure and improve forecast accuracy, and can turn predictive models into practical business solutions.

This is a technical, hands-on position within an existing machine learning team. The focus is developing and improving forecasting systems-not traditional demand planning, reporting, or dashboard development.

What You Will Own

• Design, develop, and improve demand forecasting models using Python, SQL, statistical modeling, and machine learning.
• Build SKU-level and product-level demand forecasts using historical sales, transactional, and inventory data.
• Develop time-series forecasting models that account for seasonality, trends, changing demand patterns, and business variability.
• Apply forecasting techniques such as XGBoost, LightGBM, ARIMA, SARIMA, Prophet, or comparable statistical and machine learning methods.
• Evaluate forecast performance using historical backtesting, error analysis, WAPE, MAPE, forecast bias, and model comparisons.
• Develop, deploy, and maintain forecasting models and automated data pipelines for ongoing business use.
• Work with inventory, supply chain, purchasing, and manufacturing teams to translate forecasts into better operational decisions.
• Improve forecast accuracy, inventory availability, replenishment efficiency, and production planning.
• Identify opportunities to reduce stockouts, excess inventory, manual forecasting work, and operational costs., You take ownership of complex demand forecasting problems and develop reliable Python-based forecasting solutions that improve real business operations.

Your models help the company predict product demand more accurately, optimize inventory levels, strengthen purchasing decisions, and improve manufacturing planning.

You are comfortable working independently, solving ambiguous problems, and taking forecasting models from development into practical use.

The ideal candidate is a hands-on Data Scientist who builds and improves forecasting models-not someone whose experience is limited to consuming forecasts, BI reporting, or managing planning processes.

Why Consider This Opportunity?

• Work on real-world machine learning and forecasting challenges with direct operational impact.
• Take meaningful technical ownership of forecasting models and systems.
• Collaborate directly with manufacturing, inventory, and business stakeholders.
• Join an established ML team within a growing manufacturing business.
• See the measurable impact of your work on inventory, production, and business performance.

• Professional experience in data science, machine learning, predictive modeling, or applied statistics.
• Strong hands-on Python experience personally developing time-series or demand forecasting models using real business data.
• Proficiency with Python, SQL, Pandas, NumPy, Scikit-learn, or comparable data science tools.
• Experience developing, evaluating, and improving forecasting models-not simply interpreting forecasts generated by existing software.
• Solid understanding of time-series modeling, feature engineering, historical backtesting, forecast accuracy, and model evaluation.
• Experience implementing forecasting models or repeatable predictive workflows for actual business use.
• Ability to independently investigate forecasting problems, write production-quality code, and develop practical solutions.
• Strong understanding of how predictive accuracy affects business decisions and operational performance.

Strongly Preferred

• Hands-on experience developing SKU-level, item-level, or product-location demand forecasting models.
• Forecasting experience involving inventory, replenishment, purchasing, distribution, or production planning.
• Background in manufacturing, consumer goods, retail, e-commerce, building products, or supply chain analytics.
• Experience with XGBoost, LightGBM, ARIMA/SARIMA, Prophet, Croston, hierarchical forecasting, or probabilistic forecasting.
• Experience working with large SKU portfolios, intermittent demand, seasonality, sparse data, promotions, or new-product forecasting.
• Experience deploying and maintaining production ML pipelines, including automation, performance monitoring, and retraining.
• Familiarity with AWS, Azure, GCP, Databricks, Airflow, MLflow, or similar ML engineering tools.
• Demonstrated improvements in forecast accuracy, inventory efficiency, stockout rates, service levels, or operational costs.

Location Requirement: This position is fully onsite in Newark, NJ, five days per week. Candidates must be able to commute reliably or relocate.

Compensation: $150,000-$200,000 annually, with potential flexibility for exceptional experience.

Pay: $150,000.00 - $250,000.00 per year, * 401(k)

• 401(k) matching
• Health insurance
• Paid time off