Real-Time Streaming Apps
for Enterprises. Now Made Easy.

Successful organizations need to make decisions based on data—quickly. But many organizations struggle to efficiently process massive amounts of data, rendering it less valuable over time. DataTorrent RTS 2.0 lets organizations harness the power of real-time streaming analytics with faster insights in an intuitive, scalable manner.

Sign Up for Beta

Big-Data Analytics Applications for the Non-Developer

Real-time streaming apps provide extremely valuable situational analysis, allowing data scientists and business analysts to take action in real time. Knowledge of how to create these apps was previously limited to technical developers. Project da Vinci offers an easy-to-use, web-based application builder that empowers non-developers to create real-time streaming apps.

Visual Web-Based Dashboard
for Real-Time Insights

For big data applications, true value is the ability to gather key insights in real time through a visual dashboard. Project Michelangelo features a palette of visualization tools that enables data scientists and business analysts to design a real-time dashboard on a running application.

An Enterprise-Class Platform

Scalable Ingestion

Organizations looking at streaming applications need to know their specific needs are met. It starts with being able to ingest their data type (structured and unstructured) from their data sources (data in motion and at rest) at their scale (thousands to millions of events) and at their speed (from minutes to sub-millisecond).

With more than 75 data connectors, DataTorrent RTS 2.0 enables organizations to ingest structured and unstructured data. DataTorrent RTS supports SLA driven auto-partitioning and auto-scaling of data ingestion in a fault-tolerant distributed manner across the Hadoop cluster. SLAs are met by simply setting the application property for the SLA attribute (throughput, latency, etc.) and DataTorrent RTS enforces the SLA. Data can be ingested from edge locations as well as at the data center via a hub-n-spoke architecture.

Advanced Analytics

Enterprises that are creating real-time streaming applications don't want to waste precious time re-creating the wheel for common tasks. Their focus should be on creating value-added business logic. Real-time streaming apps are built on DataTorrent RTS using a logical data flow. Data is processed as it flows through Java-based “operators.”

DataTorrent RTS 2.0 includes updates to the Malhar open-source library that provide filtering and pattern matching to simplify data analytics; large-scale multidimensional processing to compute all possibilities in real-time data; support for predictive analytics/data mining with “R” language and RapidMinor operators; and stream management such as dedup, join, select and many more.

Hadoop-Native Key-Value Data Store

Taking action based on a specific event is a key capability of a real-time streaming platform. However, an event is often more valuable in the context of other events. This is true where complex event processing (CEP) is required to look for patterns, where predictive analytics are used against multiple events to determine future actions and for creating real-time visual dashboards showing event trends.

DataTorrent RTS 2.0 includes a distributed hash map key-value data store that allows applications to store data that’s being processed in a distributed, linearly scalable fashion using hash map technology. This speeds application development and reduces cost and complexity associated with a third-party database.

Dynamic Update

Enterprise data is dynamic, changing with market and business conditions. The analytics and subsequent business decisions that an enterprise makes are dynamic as well. The ability to react instantly to changing business conditions needs to be supported by IT, not delayed due to technology limitations.

DataTorrent RTS 2.0 enables changes to an application's window time (e.g., 5-minute to a 10-minute rolling window), native A-B testing of business logic and insertion/deletion of business logic without requiring an application or system shutdown.

Real-Time Apps Made Simple

Application developers need to be experts in their organizations, and they need to be productive. What they don’t need to be is experts in distributed programming, the intricacies of the streaming application platform or Hadoop internals. DataTorrent RTS is a java-based programming environment where distributed computing and fault tolerance is provided by the underlying platform, not the developer.

DataTorrent RTS 2.0 simplifies the design and debugging of distributed big data streaming applications with the visual debugging tool. It includes the industry’s first Hadoop application packaging technology that provides a drag-n-drop application launch into a Hadoop 2.0 cluster as well as a robust management console that provides operational capabilities for the Hadoop cluster, individual nodes and containers, and the real-time streaming application.

  • Facebook
  • Twitter
  • Google Plus
  • Linkedin