Visualizations help analyzing and reason about data and are often used by us to understand the effects of our own models. But often they are not only for us. There are also visualizations that have to be provided to our customers to support them with their daily business. There are lots of great tools for building data science dashboards in the Python data visualization landscape. Most of them make it very simple for data scientists but when the dashboards have to be part of a customer facing product, other roles like frontend developers get involved. Tools that are a pleasure for one role might be a burden for the other. We will show how Altair and Vega help improving the collaboration between data scientists and frontend developers.