What is Oracle Stream Analytics ?

September 16, 2019 | 4 minute read
Thomas Vengal
Director, Product Management
Text Size 100%:

What is an Oracle Stream Analytics:

Oracle Stream Analytics (OSA) is an in-memory technology for real-time analytic computations on streaming big data. Oracle Stream Analytics can automatically process and analyze large scale real-time information by using sophisticated correlation patterns, enrichment, and machine learning algorithms. The streaming big data can originate from IoT sensors, web pipelines, log files, point-of-sale devices, ATM machines, social media, transactional databases, NoSQL databases or any other data source.  OSA offers a compelling, friendly and visually stunning real-time streaming analytics user experience for business users to dynamically create and implement instant insight solutions. The best part of this tool is that it allows users to explore real-time data through live charts, maps, visualizations, and graphically build streaming pipelines without any hand-coding. OSA executes in a scalable and highly available clustered Big Data environment, using Apache Kafka and Apache Spark Streaming integrated with Oracle’s Continuous Query Engine to address critical real-time use cases in modern enterprises.

Fig 1:  Oracle Stream Analytics highly interactive development browser UI
Industry Use-cases: 
 Few industry specific use-cases for Oracle Stream Analytics include:
  • Detecting real-time fraud based on anomaly detection machine learning algorithms.
  • Offering real-time marketing offers based on customer location and loyalty.
  • Tracking transaction losses and margins in real-time to renegotiate with vendors and suppliers.
  • Improving asset maintenance by tracking healthy operating parameters and pro-actively scheduling maintenance.
  • Improving margins by continuously tracking demand and optimizing markdowns instead of randomly lowering prices.
  • Re-adjusting prices by continuously tracking demand, inventory levels, and product sentiment on social media etc.
  • Instantly identifying shopping cart defections and improving conversion rates.
  • Upselling products and services by instantly identifying customer’s presence on the company website.
  • Improving asset utilization by tracking the average time it takes to load and unload merchandise.
  • Improving turnaround time by preparing dock and staff based on the estimated arrival time of fleet.
  • Transportation: Monitor all airline's operational events to eliminate flight delays
  • Revising schedule estimates based on actual time to enter and exit loading zones, and so on.

Technology Use-cases:

Stream Analytics can be used in the following technology scenarios:
  • Build complex event processing pipelines by blending and transforming data from disparate transactional and non-transactional sources.
  • Perform temporal analytics based on time and event windows.
  • Perform location-based analytics using built-in geo-spatial patterns.
  • Detect patterns in time-series data and execute real-time actions.
  • Build operational dashboards by visualizing processed data streams.
  • Use Machine Learning models such as regression to score current events and predict the next events.
  • Run ad-hoc queries on the results of processed data streams.


OpenSource Architecture, Big Data Ready and Enterprise-Grade Deployment
Fig 2:  Oracle Stream Analytics Integration Architecture Diagram
Stream Analytics is built on the latest Apache Spark Streaming technology and Apache Kafka infrastructure to provide full horizontal big data scale-out and 24x7 availability of mission-critical workloads. Stream Analytics starts with ingesting data from Kafka with native support of GoldenGate change data capture and automated check-pointing ensures exact-once processing and zero data loss. Distributed caching is provided through the Oracle Coherence In-memory cache that is seamlessly embedded as part of the Spark deployment. Stream Analytics can seamlessly integrate with Apache Druid and explore Druid cubes using ready to use over two dozen visualization charts. Stream Analytics lets you define a Geo-Fence and then search or correlate events with respect to the geolocation.  Machine Learning  (ML) models that were developed by data scientists using data science tools like Jupyter Notebook (Python or Spark) or R Studio or Matlab can be imported into OSA and then create your own scoring stages. This means that OSA will be your runtime for running your ML models 24x7 in production. As part of management and monitoring, Stream Analytics provides a visual representation of pipeline topology/relationships along with dataflow metrics to indicate the number of events ingested, events dropped, and throughput of each pipeline. 
Fig 3: Stream Analytics Google Maps Geofence tile layer


Streaming Patterns Library: 
Fig 4: Rich set of specialized streaming pattern library
Oracle stream analytics provides an extensive pattern library to analyze real-time event patterns for a wide array of use cases such as Real-Time Log Analytics, Digital Marketing, Predictive Maintenance, Fraud Detection, Smart Inventory, Dynamic Pricing, etc.
For example, patterns can detect correlated events, such as a door closing within 10 seconds of opening it, spikes in values, such as a temperature spike of 30% outside of the average within a given time, or detecting a W-pattern in financial trading. Patterns can also detect if a weight of manufactured items is slowly trending lower or the reading falls outside acceptable norms. 
Advanced Visualization: 

Oracle Stream Analytics includes an extensive library of 30+ visualization charts which are based on Apache Superset, which is an easy to use interface to create and explore data. A druid based cube can be created using Stream Analytics and data can be explored using interactive charts. Some of the charts are Distribution Chart, Sunburst, Word Cloud, Sankey, Time Series Chart, Treemap, Directed force Layout, Calendar Heatmap, World Map, Box Plot, Filter Box, Bubble Chart, iFrame, Bullet Chart, Streaming Chart, Table View, Big Number with Trendline, Parallel Coordinates, Markup, Big Number, Heatmap, Pivot Table etc


Fig 5: Druid Superset Cube Visualizations 
Read more about Oracle Stream Analytics

Thomas Vengal

Director, Product Management

Thomas Vengal is Director of Product Management at Oracle. Previously, he worked at Informatica as a product manager and development manager at HP Software. He has been working in the application and data management domain for over 15 years. He holds M.Tech in Data Science from BITS Pilani and an MBA from Perth Graduate School of Business.

LinkedIn Profile: https://www.linkedin.com/in/vengal/

Previous Post

Data Integration Program Guides for Oracle OpenWorld 2019

Sandrine Riley | 1 min read

Next Post

Oracle Stream Analytics 19.1 is Released

Thomas Vengal | 1 min read