Radicalbit Product suite
Radicalbit products are Fast Data systems based on true real time performances, processing events as fast as they come in.
We selected a set of state-of-the-art technologies and frameworks seamlessly integrated , working together in order to grant our customer the access to real time predictive analytics.
Machine Learning Enabled
Machine Learning is nowadays a must-have in any analytics platform. Following the evolution of A.I., the data representation moved from a mere descriptive purpose, that leaves the interpretation to specialists, to predictive and potentially prescriptive models.
With RNA, Radicalbit DS Team combined the power of machine learning with the performances of real-time streaming data processing.
Radicalbit products are machine learning enablers, systems designed to take full advantage of the possibilities granted by the use of Artificial Intelligence. We developed a library for JPMML in Flink for model serving and evaluation.
RNA is coming with a few pre-trained models ready to be deployed and refined around the customer's need.
RNA - Radicalbit Natural Analytics
Available as a separate module or as a natural compendium of RB Light, our Fast Data distribution, RNA is the module designed to provide predictive analytics. It includes machine learning capabilities and data visualization (dashboards).
- End-to-end coverage from ETL to visual streaming analytics
- True distributed real-time streaming processing engine
- Completely based on Apache Flink
- Include NSDB
- Self-service Analytics
- Notification system
- Visual pipeline / dataflow editor
- No need to code
- Predictive analytics
NSDB - Natural Series Database
NSDB is a storage solution deeply conceived having streaming real-time analytics in mind. It fits perfectly the read side of Kappa Architectures (or for systems based on Command Query Responsibility Segregation pattern). The idea it's to store metrics and to bind directly the incoming indexed data to the final users thanks to push tecnologies like WebSocket.
- View / Read oriented time series database for real-time streaming systems implementing the Kappa Architecture or the CQRS pattern
- Real-time streaming of the incoming data using websocket
- Native clustering feature
- SQL-like support
- JDBC connector
- Flink connector (sink)
- Kafka connector to ingest events / messages in Avro format
- API available for a wide range of programming languages (Scala, Java, C++, Python, Go, Ruby, C#, Node.js, Android Java, Objective-C, PHP)
- CLI
RB Light - Radicalbit fast Data Distribution
A lightweight Fast Data distribution offering the basic and best-of-breed core technologies to quickly benefit from real-time streaming and machine learning technologies.
- Open source stack based on Apache projects
- Frameworks completely integrated, customized and battle tested
- Best-of-breed fast data technologies for real-time streaming and distributed processing
- Includes Apache Flink, Apache Kafka, Apache Zeppelin, Apache ZooKeeper, Apache HDFS, Apache Ambari
- Optimized and compatibile with NSDB
- Available for both cloud and on-premise solutions
- Out of the box Kerberos support
- Integration with Flink-JPMML for evaluation and real-time serving of Machine Learning models
