Dataiku’s Data Science Studio (DSS) is an end-to-end technology product on which data analysts and scientists can create data applications from start (data plugins) to finish (deploying apps). DSS is production ready and is not only useful for insights and modelling, but also for applying predictions in real time to business operations.
DSS shortens the load-prepare-train-test cycles that are time-consuming when building predictive applications. It has enough technical and mathematical depth for data analysts and scientists to have fun cleaning, testing, and deploying data applications while remaining accessible to less technical profiles such as business analysts, marketing teams, etc. It allows data scientists, developers, and business analysts to work collaboratively. DSS allows them to easily show their work to outlying teams thanks to drag and drop graph visualization technology, visual flow charts, project pin boards, etc.
Data Science Studio’s technology uniquely combines visual transformation, coding capabilities, and machine-learning so that end-users can interactively design data transformations and build predictive models at scale. DSS can connect, load, and execute processes on remote data stores (NoSQL), clusters (Hadoop) or specialised machine learning clusters. DSS automatically analyses data, chooses, and tries several feature transformations, compares and combines core algorithms to create a unique optimised algorithm for the unique customer data and creates data-driven workflows that combine SQL, Python, Hadoop programming, and interactive charts. It also provides the ability to recompute only what’s needed in complex production workflows and automates recovery from hardware crash and missing data.