Customer intelligence platform for finance

We explore how deep learning methods can be used to analyze sequences of customer interactions and how the insights gained from such analyses can be used for spend optimization. We gradually build a solution that can be applied to several common scenarios, including campaign optimization, channel optimization, and channel parameters optimization.

In this article, we focus on the learning of useful semantic representations (embeddings) for products and customers using neural networks. These embedding vectors then can be used to improve customer analytics and segmentation as well as features in the downstream customer intelligence software and components such as propensity and recommendation models.

In this tutorial, we discuss how traditional targeting and personalization models such as look-alike and collaborative filtering can be combined with reinforcement learning to optimize multi-step marketing action policies (aka Next Best Action policies). We use methods that were originally developed and tested by Adobe and Alibaba, and they proved to be effective in a number of practical use cases.