The Client needed a solution that will help optimize its marketing expenses on product promotions, assist the sales team with planning their future activities and demand estimates.
Solution:
Sales prediction allows companies to spot potential issues or risks and design appropriate corrective actions to mitigate them. In essence, sales forecasts help companies with sales planning, demand planning, inventory control, financial planning, internal control, gain insights, and marketing benefits.
The project aimed to create a solution that will allow quick and timely access to statistical models based on the client’s datasets to solve the specified problems, e.g. predicting sales, demand, and simulating marketing strategies to increase the net volume of sales.
We use a time series regression analysis approach and utilize conventional ARIMA models as well as modified LSTM deep learning network. Such approaches help with getting valuable insights on sales drivers and provide the client with predictions, while being interpretable.
We have also constructed a dashboard, which automatically updates models and forecasts based on the new data being uploaded.
Technology stack:
Django, Keras, Tensorflow, Plotly, Dask.
