Zerve Reviews

Pros

Before we started using it, my data science team was spending around 80% of its time dealing with downstream engineering work which left very little room for actual model development and R&D. Since adopting it, that engineering burden has been almost completely removed and our cycle times have improved a lot as a result. The collaborative coding environment also helps my data science team work much more effectively with other teams. Another thing I really appreciate is the interoperability between Python and SQL and it's a plus that R is supported as well.

Cons

Some MLOps capabilities such as model monitoring, aren't built in yet. That said, the team has said those features are currently being developed and for now we're using an MLflow integration as a temporary solution.

Pros

It has really organized the way we work and helped us move into production much faster. A good example is our R&D team, they used to share research through a bunch of different methods but now we create an API endpoint that can be used across the whole organization.

Cons

I'd love to see support for building native real-time apps.