useR!2019 Toulouse recap

[This article was first published on r – Appsilon Data Science | End­ to­ End Data Science Solutions, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Conferences like userR & EARL are the R events to attend every year and personally, and as a company, I can’t imagine skipping one. It’s an important place to be if you want to be up-to-date with the R technology and build up your presence in the community. Our team have given rave reviews after speaking at past useR events in Brussels or Brisbane.

This year the conference was hosted by the great city of Toulouse. I hope that I won’t offend locals by saying that it resembled Sevilla with all of the people sitting next to the river sipping wine, the heat, and friendly atmosphere. I regret that I couldn’t take my family to stay for the weekend in the city.

The conference was packed! With over 1200 attendees, six parallel tracks and a constant fear of missing out. Every minute you face a decision of which great talk to listen or with whom to take a coffee having so many great individuals in one place. Though It doesn’t spoil the unique atmosphere of the R community – everyone is super friendly and open. The R ecosystem is a mix of folks of different backgrounds.  Being exposed in the past to many different tech communities, I find it truly unique. The gala dinner on Wednesday took place at the “Cité de l’Espace”; it was a great opportunity to rest after the whole day and get to know other community members.


Each and every keynote speaker impressed me with distinct values. However the most inspiring was the presentation by Julien Cornbise. Julien spoke on topics that are the most important for our team — applying AI to do good for humanity. Although it’s fairly easy to generate plenty of ideas that could revolutionise the world, the implementation & building a real, viable case creates a constant uphill battle. At Appsilon Data Science, we are exploring the ways we can contribute in a positive way with our machine learning (ML) skills to fight global problems.


Most memorable talks for me: 

  • Machine Learning Infrastructure at Netflix by Savin Goyal (slides)
  • Measuring inequalities from space. Analysis of satellite raster images with R by Piotr Wójcik (slides)
  • Building and Benchmarking Automatic Machine Learning Systems by Erin Ledell (slides)
  • Authentication and authorization in plumber with the sealr package by Friedrike Preu (slides)


My presentation

I had a great opportunity to be one of the speakers. I gave a talk on best practices for building Shiny enterprise applications. As a company we build and deliver data products for enterprise. As more than half of our commercial projects are done with R, and our core competency is building decision support systems with Shiny, we wanted to share our best practises.

Shiny has unmatched speed of development. I remember that the first deal I won with a fortune 500 company was thanks to implementing the MVP of the application in 24 hours. This is something I am not able to repeat at this level with any other framework or language.

When compared to spreadsheets,  R Shiny offers more beautiful interface and automation. You can automate anything in the background and can easily implement shared resources, scalability, reproducibility, and most importantly the source code.

Source code is also the main benefit when compared to BI, but next to that you get almost no running fees, full customization and better ML possibilities.

Let’s say you have a data product to build. If you need more than a spreadsheet or a BI dashboard, but you are not 100% sure of the scope – Shiny is a perfect choice.

My presentation slides can be found here.


Questions I have received:

  • linting – when to do it and how ← I advised to do it as often as possible. Plug it in to your IDE, plug it in to you Git pre push hook, plug it in to your continuous integration. Just make sure to plug in the same rules everywhere!
  • integration tests ← test larger logic than in unit tests, use mocks when relevant and agree with your team what is an integration test and what is not.

Last word & some photos

All in all, I enjoyed the useR quite a bit. A big shout out to the organizers! I can’t wait for next year’s edition! In the meantime you can catch our team members at the Why R conference in Warsaw (26-29 September), where we have our headquarters.

Event link:

Article useR!2019 Toulouse recap comes from Appsilon Data Science | End­ to­ End Data Science Solutions.

To leave a comment for the author, please follow the link and comment on their blog: r – Appsilon Data Science | End­ to­ End Data Science Solutions. offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)