Forecasting Pricing

Manufacturing ‱ Artificial Intelligence

Challenge

Our client’s product prices have high variability and are very dependent on prices in the United States.

Weather anomalies such as storms can have an important impact on our client’s products.

Current price forecasts do not consider weather as a variable.

Solution

Built a statistical model to forecast the production of their product by region.

Built a pipeline to ingest weather data from across Canada and linked it to the model to improve accuracy.

Results

Improved the accuracy of price forecasts.

Better understanding of impact of weather anomalies on pricing.

Leveraged weather data to improve the variance in product pricing.