Striim ingests real-time data from transactional databases, log files, message queues, and sensors. For enterprise databases, including Oracle, Microsoft SQL Server, MySQL, and HPE NonStop, Striim offers a non-intrusive change data capture (CDC) feature to ensure real-time data integration has minimal impact on source systems and optimizes the network utilization by moving only the change data.
In-Flight Data Processing
As data volumes continue to grow, having the ability to filter out and aggregate the data before analytics becomes a key way to manage the limited storage resources. Striim enables in-flight data filtering
and aggregation before it delivers to Hadoop to reduce data storage footprint. By performing in-line transformation (such as denormalization) and enrichment with static or dynamically changing data in memory, Striim feeds large data volumes in the right format without introducing latency.
Striim is designed to meet the needs of mission-critical environments with end-to-end security and reliability — including out-of-the-box exactly once processing — high-performance, and scalability. Users can focus on the application logic knowing that from ingestion to alerting and delivery, the platform is bulletproof to support the business as required.
Fast Time to Market
Intuitive development experience with drag-and-drop UI along with prebuilt data flows for multiple Hadoop targets from popular sources allow fast deployment. Striim uses an SQL-based language that requires no special skills to develop or modify streaming applications.
Operationalizing Machine Learning
Striim can pre-process and extract features suitable for machine learning before continually delivering training files to Hadoop. Once data scientists build their models using Hadoop technologies, these can be brought into Striim, using the new open processor component, so real-time insights can guide operational decision making and truly transform the business. Striim can also monitor model fitness and trigger retraining of models for full automation.