How is Workshop different from Slate in Foundry?

Question

I see that there are Slate and Workshop in the Foundry platform. May I know real business cases where we can use Workshop and Slate? How are these different and where can these best fit? can anyone shed some light on this?

Answer 1

Generally, your first choice should be to reach for workshop. Workshop is much more "on rails" than slate, which means the resulting applications are much more maintainable and easier to understand and customize.

Workshop can be thought of as a flexible ontology-based "interactive dashboard maker", where you position and glue together widgets (possibly with some glue code), all of which display or interact with ontological objects (i.e. the objects you see in object explorer.)

Slate is a much more general tool, and you can think of slate a little more like a WYSIWYG HTML designer. It allows you to make fairly arbitrary pages that query arbitrary APIs and data sources. This sounds more powerful, but it's also a much more complex model and things can get messy very quickly.

Workshop is the newer tool, and in a way is better integrated with the foundry platform -- since it interacts with objects (ie allows you to let the user edit objects), knowledge is captured directly and with no delay in the ontology, and other users can directly benefit from it. Since it's integrated with the ontology, it also understands how to meaningfully present data, rather than just showing the users rows and columns.

This means with workshop you can realize quite a few workflows without having to write any code at all, such as inbox-style triaging workflows (alerts / work-items being surfaced in a users inbox, who can then send them on to other users or close them out, etc.) Here's a quick example you could set up in probably less than half an hour, assuming the data is in place:

Slate is a much older tool (it predates foundry) which makes it more powerful in some ways, but if you want to enable users to capture knowledge in slate, you will need to do something more custom like write back to a postgresql database or similar, so it's a lot less "native".

Hence my recommendation is to always consider using workshop first. If you have a workflow that's almost possible to implement in workshop, but there's just some small missing feature, I'd recommend to talk to your palantir contact, as the workshop development team would likely be interested in enabling this workflow in workshop.