Welcome to Kamanja’s documentation!

Welcome to Kamanja’s documentation!

Kamanja is a high-performance real-time event processing engine. The engine provides a run-time execution environment to run programs to compute analytics and KPIs, as well as models encapsulating machine learning methods. It provides adapters to connect to both synchronous and asynchronous data sources. Kamanja is built ground-up as a distributed processing engine that can be deployed in a cluster mode to handle large data volumes and complex model computations.

Kamanja is open sourced under the Apache™ license. It includes:

  • Real-time computation engine
  • Input, output, and storage adapters
  • Support for Java, Python and Scala as model development languages
  • Model run-time based on Directed Acyclic Graphs (DAG)
  • Support for PMML
  • Metrics capture across all Kamanja components
  • Support for serialization and deserialization of data using JSON, CSV, or KV formats
  • Out-of-the-box support for cluster mode for a scalable and fault-tolerant processing environment
  • search and resource management REST APIs and pre-built connectors
  • Out-of-the-box integration with Apache™ Hadoop® and data stores such as Apache HBase™ or Apache Cassandra™
  • Out-of-the-box integration with messaging systems such as Apache Kafka or IBM® MQ.

For a more detailed description of Kamanja, see What is Kamanja?.

How to use this documentation

  • Ligapedia is a set of free-standing articles, arranged alphabetically, about terms and concepts related to using Kamanja. Articles tell you waht the term means and include references to other documents (within the Kamanja doc set and elsewhere) that give more information. Other documents often link to these articles when using specialized terms

  • Planning and installing Kamanja gives instructions for installing a Kamanja cluster. For demonstrations and development, you can install a single-node cluster; for production systems, you need to install a multi-node cluster.

  • Working with Models describes the structure of Kamanja applications and how to create and implement the analytical models, messages, and containers that make up an application.

    Specialized guides are provided for implementing applications using Java, Scala, PMML (which is also used to implement trained R models), JTM, and Python.

  • Working with Adapters gives instructions and examples of writing the adapters used to pull data into the Kamanja environment, export the results to another application, or store data in the Kamanja factory.

  • Kamanja Architecture discusses the Kamanja architecture and components.

  • Reference pages provide detailed technical reference material about the following:

Machine Learning Guides

Machine Learning Guides

Administration Guides

Administration Guides

Indices and tables