Skip to main content

Introduction

What is Nussknacker?#

Nussknacker is an event processing and decision-making platform that allows designing real-time decision algorithms. It is capable of high throughput and low latency event filtering, transformation, enrichment, correlation and aggregation. It can handle any data - call detail records, financial transactions, clickstreams, network traffic metadata, IoT-generated payloads, images, logs, etc. The platform is built on top of and integrates with industry-standard components, e.g. Apache Flink and Apache Kafka.

Why Nussknacker?#

Real-time data decisioning quickly becomes one of the key ingredients of marketing, monitoring, IoT or fraud detection. The ability to ingest large amounts of data within seconds, process them and take actions can become a real competitive advantage.

Development of real-time analytics and decisioning applications is not a trivial task - in-depth technical knowledge is needed, e.g. of Java, Flink and Kafka. Many companies do not have large enough development teams to define decision logic with code. Furthermore, most use cases require decision algorithms that would be best defined or changed by not-so-technical users - analysts or business people.

This is where Nussknacker kicks in.

An essential part of Nussknacker is a visual design tool for event processing. Decision scenarios can be authored from predefined processing components using simple drag and drop. Components range from simple ones like filters and switches to complex ones, allowing aggregations in different types of time windows. Once authored, with a click of a button, scenarios are deployed for execution to Flink - one of the most advanced platforms for stream processing.

Nussknacker allows knowledge workers to harness the power of stream data processing platforms like Flink without coding low-level stream processing logic. Users who author scenarios do not need to have in-depth knowledge of Flink and its API. Moreover, the ideate -> author -> test -> deploy cycles can be very short, allowing to quickly test hypotheses about data and react to changing business or technical needs.

Nussknacker in action

Scenario authoring with Nussknacker