SaaS platform which allows data scientists to setup experiments, visualise results and make educated decisions on their products.
An animated identity that acted as the platforms brand and also loading transition.
To fully grasp how a data scientist would perform an experiment I studied their work flow and built up a framework based on requirements.
The best way to understand an experiment is in the form of a lifecycle, of 4 key stages:
Create, monitor, author & research.
An experiment would usually be setup by a data scientist as there are several specialised stages to run calculations on how the experiment should be run.
Once an experiment is live and data is being collected the view will update to show health metrics, allowing the user to see if there are any anomalies and also draw early conclusions.
After an experiment has completed, teams can make a decision on whether to keep the experiment live or alter it. This information is recorded and can be used to instigate further research on a product area.
Data scientists now have the ability to show performance of their experiments in real time.
Experiment data being presented to the business.