Eastwind DataFlow

Main article: Definition of Business Intelligence

The Eastwind DataFlow platform automates the process of managing data processing on a Hadoop cluster. Allows you to create and manage import, export and calculation tasks on a distributed computing Hadoop cluster.

2022: DataFlow Capabilities

The Eastwind DataFlow platform has an intuitive interface in which date scientist it can explore data, AIML develop/models and run them into a production on the fly without wasting time finding the right ecosystem tools. big data

DataFlow optimizes the current infrastructure through a set of ready-made solutions for loading, calculating and uploading data, provides interaction with Hadoop utilities.

According to information for December 2022, the platform allows:

• Integrate all types of DataSources
• Preprocessing data
• Create and populate DataLake
• Select and analyze data
• Create analytical predictive models
• transfer the research results to the production system of the Hadoop cluster;
• automatically monitor the quality of work and metrics of the created models;
• Optimize and adjust models.

2018: Module announcement

On June 21, 2018, Eastwind introduced the DataFlow module for working with analytical models on the Hadoop cluster.

EW DataFlow helps analytics work with data on Hadoop, bypassing DevOps engineers. The module connects directly to the cluster and outputs all the necessary information about the available data to a convenient UI. Thus, EW DataFlow acts as an adapter or adapter for Hadoop.

In a comfortable and intuitive environment of the module, a data scientist can work with big data on familiar tools: quickly and without intermediaries. Developers only need to deploy the system. Under the hood of the module are cluster tools for calculations, but the analyst will write all the code directly in UI in python.

Previously, when there were problems with data on Hadoop, two people sat down to work on one task: a data scientist and a developer, "said Pavel Olifer, head of social analytics at Eastwind. "The company was wasting time and money. We created EW DataFlow so that this would not happen. The module makes the work of the data scientist on the Hadoop cluster transparent. He wrote the code himself, launched it himself, you monitor it yourself. If anything, he corrected it himself. After all, business analytics should be fast and relevant. Only then will it give the desired effect and bring profit

The DataFlow module provides a single environment for working with analytical models on a cluster. In an intuitive UI module, a data scientist will be able to investigate data, develop mathematical models and algorithms and run them into a product right on the Hadoop cluster.

With EW Dataflow, you can:

• Connect new data sources.
• Perform any data processing (sampling, research, model building, monitoring, etc.).
• Run models in the product and tune them.
• Instantly learn about problems in work, find errors in the code and edit.
• Manage all calculations on the cluster.
• Export Work Results to Files

You can load exported data files into any analytical systems.

EW DataFlow Module Delivery Options:

• As a separate product - for those who already work with data automatically or manually,
• With the EW Social Analytics platform - for those who need a comprehensive analytics solution.