Automated metrics calculation in a dynamic heterogeneous environment
Authors:
Craig Boucher,
Ulf Knoblich,
Daniel Miller,
Sasha Patotski,
Amin Saied,
Venky Venkateshaiah
Abstract:
A consistent theme in software experimentation at Microsoft has been solving problems of experimentation at scale for a diverse set of products. Running experiments at scale (i.e., many experiments on many users) has become state of the art across the industry. However, providing a single platform that allows software experimentation in a highly heterogenous and constantly evolving ecosystem remai…
▽ More
A consistent theme in software experimentation at Microsoft has been solving problems of experimentation at scale for a diverse set of products. Running experiments at scale (i.e., many experiments on many users) has become state of the art across the industry. However, providing a single platform that allows software experimentation in a highly heterogenous and constantly evolving ecosystem remains a challenge. In our case, heterogeneity spans multiple dimensions. First, we need to support experimentation for many types of products: websites, search engines, mobile apps, operating systems, cloud services and others. Second, due to the diversity of the products and teams using our platform, it needs to be flexible enough to analyze data in multiple compute fabrics (e.g. Spark, Azure Data Explorer), with a way to easily add support for new fabrics if needed. Third, one of the main factors in facilitating growth of experimentation culture in an organization is to democratize metric definition and analysis processes. To achieve that, our system needs to be simple enough to be used not only by data scientists, but also engineers, product managers and sales teams. Finally, different personas might need to use the platform for different types of analyses, e.g. dashboards or experiment analysis, and the platform should be flexible enough to accommodate that. This paper presents our solution to the problems of heterogeneity listed above.
△ Less
Submitted 2 December, 2019;
originally announced December 2019.
