Supercharged Customer Service with Machine Learning

In this blog post, we will simulate a real-world customer service use case and use tools machine learning tools of the Hugging Face ecosystem to address it.

We strongly recommend using this notebook as a template/example to solve your real-world use case.

Defining Task, Dataset & Model

Before jumping into the actual coding part, it's important to have a clear definition of the use case that you would like to automate or partly automate. A clear definition of the use case helps identify the most suitable task, dataset to use, and model to apply for your use case.

Defining your NLP task

Alright, let's dive into a hypothetical problem we wish to solve using models of natural language processing models. Let's assume we are selling a product and our customer support team receives thousands of messages including feedback, complaints, and questions which ideally should all be answered.

Quickly, it becomes obvious that customer support is by no means able to reply to every message. Thus, we decide to only respond to the most unsatisfied customers and aim to answer 100% of those messages, as these are likely the most urgent compared to the other neutral and positive messages.

Assuming that a) messages of very unsatisfied customers represent only a fraction of all messages and b) that we can filter out unsatisfied messages in an automated way, customer support should be able to reach this goal.

To filter out unsatisfied messages in an automated way, we plan on applying natural language processing technologies.

The first step is to map our use case - filtering out unsatisfied messages - to a machine learning task.

The tasks page on the Hugging Face Hub is a great place to get started to see which task best fits a given scenario. Each task has a detailed description and potential use cases.

Finding suitable datasets

Since we consider the hypothetical use case of filtering out unsatisfied messages, let's look into what datasets are available.

In addition, the Hugging Face Hub offers:

Output: