Predicting Balance Sheet items and Cash Flow is a major challenge for today's CFOs around the world, and it became exceptionally important during a cataclysmic event like the Coronavirus pandemic. They have been struggling with such topics for years, and most of the implementations we've seen are something very naïve and usually time-series autocorrelations.

In this session I will show how to approach such a project (setup, tools, methodology, algorithms), what are the common obstacles you need to surpass (data integrity and format, interpreting data, choosing the algorithms, how to treat outliers), how to run models training at scale using AutoML, and when and what to customize for better prediction.

This presentation is based on the work we've done for a few large enterprise customers (Oil & Gas and other industries), helping them predict Balance Sheet items and Cash Flow.