Competition in the ‘Fresh’ category in food and grocery in India is Tremendously high. To run a viable business, More needs to simultaneously manage the in-stock availability of fresh produce while minimizing wastage. To balance these competing priorities, We needed to build a very granular forecast at the store-item-day level. We prioritize the development effort based on the ABC-XYZ framework that is, The store-item combinations will be plotted on a 3x3 matrix: ABC axis of sales saliency (A – high, B-Medium, C-low) and XYZ axis of forecastability (X-easier to forecast, Z-difficult to forecast) based on the historical pattern. The forecast accuracy of items in ABC-XY buckets will be much superior to the Z bucket. However, for combinations in the Z bucket, Deepsight significantly outperformed traditional methods like exponential smoothing yielding an incremental 10% forecast accuracy. This was possible because of Deepsight’s ability to learn other SKUs (XY) patterns and apply them to highly volatile items in the Z bucket. Using Deepsight, you will able to increase your forecasting accuracy by 3x, reduce wastage by about 20% for the fresh produce category. Deepsight provides a distribution of forecasts which helps you to optimize yours under and over forecasting costs leading to stock-outs at 3% and improved gross margins. This makes it easier for your store managers to place more accurate purchases orders by looking at the daily forecasts. You can also expand the model to other categories, iterating with additional related datasets, and adding newer data to Deepsight to continuously improve the model accuracy.”