Explainable Artificial Intelligence

XAI is hot. But XAI is hard.

Ever wondered why you are recommended a particular song or playlist? Or, why your favourite news provider recommends a particular article? Or, how your online retailer is able to suggest the next product you should buy?

Explainable Artificial Intelligence (XAI) is an emerging field of AI that offers the ability to answer these questions and address how black box decisions of AI systems are made, for both the customer and the business.

While the topic of XAI is hot, XAI is equally as hard. So, how can business go beyond the buzzword?

In this two-part series, our lead Data Scientists details how to get started.

Watch Now

Part I: When does your organization need to think about XAI?

Datatonic’s Senior Data Scientist, June He, walks through when your organization needs to think about XAI, the “why” and “how”, and defining the purpose of XAI for your business.

Part II: A Data Scientist’s starter guide to doing meaningful XAI

Datatonic’s Data Scientist, Laurence Moscrop, provides a hands-on guide to tackling an ML problem from beginning to end where explainability is an important goal. He offers a starting point for you to launch further into this evolving field through five steps.

Case Study: Real-Time Recommendations and XAI for NewsUK

NewsUK aspired to personalize the reading experience to their millions of readers, but previous experiments had left the team frustrated.

They needed a recommender system that was high-performing, able to process large volumes of data in real-time and cost-effectively, and able to provide explainability for each recommendation being served.

Datatonic has developed a state-of-the-art recommender system that personalized articles in real-time for The Sun, The Times and The Sunday Times, with Explainable AI as a key component in the model design.

Read more about our case study here.