Machine Learning Engineer

We are looking for a Machine Learning Engineer who has built predictive models around time-dependent data and deployed them into real production environments.

The primary focus is forecasting and predictive analytics used to improve decisions around demand, sales, inventory, purchasing, and operational planning.

This is not an LLM or generative AI role. We are specifically looking for experience building predictive models where time, historical patterns, seasonality, trends, and future outcomes matter.

What You Will Own

• Build and improve production-grade forecasting and predictive models using time-dependent data.
• Develop SKU-level, demand, sales, inventory, or other operational forecasting systems.
• Own models from development through deployment, monitoring, retraining, and ongoing improvement.
• Build and optimize data pipelines supporting predictive ML systems.
• Measure model performance and continuously improve forecast accuracy.
• Partner with business and operational teams to turn model outputs into better decisions.

What Success Looks Like

• Forecasting accuracy improves measurably over time.
• Predictive outputs are trusted and actively used by business teams.
• Models perform reliably in production at meaningful data scale.
• Forecasts improve purchasing, inventory, capacity, or operational decision-making.
• Models can be iterated and deployed quickly as business conditions change.

• Proven experience building time series, temporal, forecasting, or other time-dependent predictive models.
• Experience deploying machine learning models into production.
• Strong Python and machine learning/statistical modeling skills.
• Experience working with large transactional or operational datasets.
• Experience building or supporting production ML data pipelines.
• Ability to connect model performance to real business outcomes.

Especially Relevant Experience

Experience predicting outcomes such as:

• Demand
• Sales
• Inventory requirements
• Purchasing needs
• Capacity
• Production volume
• Customer behavior over time
• Other future operational or business outcomes

Experience with methods such as gradient boosting, regression, tree-based models, classical statistical forecasting, probabilistic forecasting, or deep-learning approaches to temporal data is relevant.

The important requirement is not a specific algorithm. It is evidence that you have successfully built predictive models where time-dependent data was central to the problem.