Dataiku will release a version 7 of its platform. The update adds functionalities to promote collaboration between data scientists. It also provides additional insight into the impact of enterprise artificial intelligence (AI).
The platform now has improved Git integration, allowing data scientists to create, remove, and push and pull Git branches directly from Dataiku. Users can easily duplicate projects to make changes in a test environment without impacting the project. Once the changes are completed, they can be implemented in the original project (the changes are documented in Git).
Additionally, in this version, Dataiku works with what is also called ‘explainable AI‘. Originally, models don’t provide insight into why and how they come to certain conclusions. This makes it difficult for the business to explain why decisions are made based on the models. So now an explanation is provided by describing which characteristics or elements have the greatest impact of the model. Version 7, for example, offers interactive visualisations to explain the prediction.
Another interesting update provides more flexibility with Kubernetes. It builds on the possibilities for managed Kubernetes clusters by running web applications on the clusters. This creates space for more simultaneous users and provides a fast and flexible back-end for demanding AI environments.
A few months ago, Dataiku launched version 6. At that time, there was also a strong focus on the easy deployment and management of Kubernetes clusters.