One of the big strengths of the CodeSignal platform is the consistency and simplicity of the UI; the theme and layout can be customized to the candidate’s needs, and regardless of the task type, users know where to find the description, tests, and coding area. Furthermore, experienced devs will find the coding experience familiar, as they’re able to select from classic text editors such as Vim, Sublime text, and Emacs.
In this episode, we look at a data analysis task, which is quite different from previous types of tasks, but despite this, the appearance is familiar and inviting. In order to solve this task, the candidate would need to interact with a database and an API using a programming language (rather than simply writing queries, as we’ve seen in previous tasks). The idea is to more accurately emulate the workflow of a backend developer, who may need to import various libraries, and pull in data from diverse sources.
To avoid usage limitations or other potential complications, the simulated API lives on the CodeSignal server, but the candidate can interact with it as they would with a real API. It can also be configured to return errors occasionally, to ensure that the candidate’s code is robust enough to handle these instances. Finally, because Docker is used to manage the coderunner server, the simulated API and database can be setup and destroyed very quickly.