Slides have now been posted for many of the talks given at the recent Effective Applications of the R Language (London) conference, and I thought I'd highlight a few that featured Microsoft R.
Chris Cole manages the deployment of R at Investec, supporting investment and risk teams worldwide. Despite some initial grumbling about having to use "old" versions of software, the R developers ultimately appreciated the improvements in collaboration and IT support derived from a standardized "R Suite" company-wide.
Investec also supports R deployment for other analysts who don't necessarily use R, by integrating R into desktop applications via a centralized DeployR server. This provides managed R-based analytics throughout the company, without the need to manage R on individual desktops.
Adam Rich has been evaluating the in-database capabilities of SQL Server 2016 R Services for Beazley Group. In his presentation, he compared and benchmarked various architectures and processes for applying R to data in SQL Server. The results varied by the type of load applied (calculating a mean of 60,000 rows vs 6,000,000 rows); pulling data from SQL to R was faster for the small loads, but pushing R code to the database was better for heavy loads.
Ben Downe, head of business intelligence at British Car Auctions, deploys R models developed in-house to the production environment using Azure ML. He describes the benefits of using the "one button API" to publish R functions as a Web service as "simple creation and destruction of new containers for new models / new test versions", "doesn't need to use IT resources", and "value for money". He also described using the Custom R Modules feature of Azure ML to simplify the process of updating R functions in deployment.
You can also check out the slides from my own keynote presentation, The R Ecosystem. There, I suggest why R is such a popular language despite being a language designed for data science (rather than a general purpose language), and having some idiosyncratic features (which seem quirky to some programmers, but make it easier to use for those with a background in data science). In addition to the language itself, there's a lot of value in the ecosystem surrounding R: the community resources, governing bodies, and companies (including Microsoft) who provide services and software to make using R more effective.
For more examples of R applications from the EARL London conference, check out the list of presentations (with slides) at the link.
EARL London: Speakers