WHAT IS DATAFLOW?

DataFlow — all-in-one platform for data processing. Track the work from the moment of data acquisition up to the implementation of machine learning models.
DataFlow helps managing calculations on the Hadoop-cluster like on the local machine: easily and quickly.

Use familiar Jupyter and Zeppelin notebook interfaces and implement the results of machine learning to your business processes in a few clicks.
Get the full power of Big Data tools in one intuitive and modern Web interface for better productivity.

DataFlow UI

FEATURES

DataFlow speeds up the core tasks of Data-Scientist

Import and export

Get control over incoming and outgoing data streams of your Hadoop-cluster in one click. Aggregate data from any sources. Integrate the results of your work with any system

Modeling

Create predictive models and control calculations on the Hadoop-cluster. Visualizing of your calculations interrelation gives the full picture of your data usage

Research

Work with data in familiar Jupyter interface and transfer your calculations from the "Notebook" to Hadoop -cluster by yourself, without data-engineers

Monitoring

Monitor data-processing cycles. Control the quality of machine learning, track data availability 27/7, quickly react to any changes and make smart desisions

LEt Dataflow Solve all your problems

Data-engineers, DevOps and developers have their technology stack such as
Java, Scala, own Scheduler, etc. Using these tools they bring the models of Data-scientists into production.

Data-scientists use Python with its excellent libraries.
After creating a model on the local machine, it should be transferred to the Hadoop-cluster.

COMMON PROBLEMS

Data Loss

Data-scientists cannot access the necessary data directly, understand how data is stored on the cluster and be aware about the type of data which is currently available

Loss of Accuracy

Accuracy of calculations was lost during the transfer from Python to Java/Scala. KPI was reached on the local machine, but was unpleasantly far away from the one on the cluster

Waste of Time

Difficulties in communication between data-scientists and data-engineers; long time-to-market, and other difficulties of modeling without DataFlow

DATAFLOW: WHY and WHAT FOR?

We wanted to make the life of each Data-scientist happier and developed DataFlow!

Difficulties

For many years we use Hadoop.
And every day we faced the same routine tasks. Performance of the tasks required interaction of many people

Stress

We wasted time, money and were eager to stop routine work. That's why we started automatization and created many amazing tools.

Solution

Then we developed very intuitive and user-friendly interface for the tools. By that we started the era of DataFlow – easy-to-use modeling environment

IMPROVE EFFICIENCY OF YOUR TEAM