Machine Learning

Anticipate. Predict.
Create. Your Control.

Kx AutoML helps users to automate the process of applying machine learning techniques to real-world problems and opens up machine learning capabilities to a wider audience. The AutoML framework provides a complete machine learning workflow, including pre-processing, feature engineering, model selection/optimization and report generation. Follow this link to see how it can be used to solve regression problems using the FRESH algorithm.

Intel’s Optane DC Persistent Memory

Benefiting from the latest
developments in Intel hardware

Intel’s Optane Persistent Memory enables on-disk databases to run faster and in-memory databases to scale. In this release Kx extends its support for Intel’s Optane DC persistent memory using App Direct Memory mode, giving you greater control over the DRAM/Optane allocation of data. Kx ultra high-speed performance – the fastest just got faster.

Cloud to Edge Deployment

Run Anywhere

Kx Streaming Analytics’ small footprint and single-stack architecture enable it to be deployed seamlessly and identically across multiple environments. Running the same software to solve complex business problems, whether on-premises, in the cloud, at the edge, or in any hybrid configuration of the three, makes it quicker to deploy, easier to maintain and cheaper to operate. It supports multiple cloud architectures including serverless deployments that enable developers and organizations to focus on delivering business results rather than worrying about hardware and infrastructure requirements.

System Resilience and Reliability

High-Availability and Fault Tolerance

Data doesn’t stop, so the decisions and automation that depend on it cannot stop either. That’s why we support high-availability cluster configurations with node-level quorum capabilities that provide resilience in the face of network failures. Advanced monitoring of processes, applications and nodes along with a range of hot/warm/cold configurations and data replication options cater for different cost/risk profiles. Fault tolerance is not fault elimination, however, so to cater for the most demanding, mission-critical environments we have extended capabilities around continuous data ingestion, storage and advanced query routing to ensure no downtime, no data loss or compromise in data integrity in the case of disruption.