A scalable and robust data science lab and collaboration platform
February 22, 2021

A scalable and robust data science lab and collaboration platform

Edgar Bahilo Rodríguez | TrustRadius Reviewer
Score 9 out of 10
Vetted Review
Verified User

Overall Satisfaction with RStudio

We are using RStudio Teams as central lab and prototyping environment for our Data Scientists. We use RStudio Connect to share plots, prototype dashboards and APIs with our colleagues, stakeholders and software developers.

RStudio is part of our internal data science platform and it is being used across product lines within our division (Industrial Applications).

RStudio Teams addressed the problem of having a prototyping environment for everyone. Users can decide which IDE would like (JupyterLab, RStudio, VScode) instead of forcing them to go with the default classical cloud Jupyter environment.
  • Close to open source.
  • Kubernetes integration.
  • Run on-premises.
  • Cloud agnostic.
  • A bit hard to set up.
  • E2E Data Science Platform.
  • Fast prototyping.
  • Collaboration.
We feel that RStudio Teams is so far one of the best prototyping environments for data scientists. It is much more robust than standard JupyterLab/Jupyter Notebook instances in the cloud and it supports better authentication methods, allows to share your content via RStudio connect in multiple ways (plots, dashboards, pins etc...) and it also has some internal load balancing mechanism that solves some of the problems that we have with Jupyter. It also allows us to lock and configure different user profiles in a more easy manner.

In comparison with SageMaker, we find the Kubernetes Sessions cheaper than the interactive development in SageMaker. We also like the fact that we are not forced to a JupyterLab based environment. On the other hand we miss some good features of SageMaker like the possibility of seeing your feature groups, model registry, training jobs, training logs and have nice autoML. We use both of them daily and we are very happy.

In comparison with DataRobot, we feel that RStudio teams has a much more reasonable pricing strategy. It does not offer all of the nice UI and autoML capabilities but we fill that we are covered to what we have now.
  • Centralized Prototyping Environment.
  • Self-service shiny deployment with minimal ops required.
  • Self-service API deployment with minimal ops required.
  • Matching Python and R users in the same place.
  • Separation between interactive sessions and on demand jobs.