Real-time analytics makes operational data available to users or dependent processes as soon as it becomes available to the database. Real-time analytics enables a business to make instantaneous decisions to adapt to changing market conditions to maintain its competitive edge.
Why Use Real-Time Analytics?
Before database technology could ingest and analyze data in real-time, businesses had to rely on pre-aggregated data produced by batch processing systems, resulting in decisions made on stale data, missed opportunities, and lower decision-making confidence. The time it takes to get to data and analyze that data directly impact business agility.
Using Streamed Data for Real-Time Business Intelligence
Event based analytics ideally complement real-time analytics. Streaming data platforms such as Apache Kafka can capture events such as updates to stock prices and allow applications and databases to consume these changes. Using a data integration solution can capture the Kafka event message, which can then be inserted into a database table. This database table, in turn, can be the data source for a tile in a visual BI dashboard plotting the changes in the stock price in real-time.
Real-Time Data Trumps Stale Data
All categories of analytics can benefit from the most current data. Below are some examples:
- Cognitive analytics augment human decision-making by tapping into machine learning, semantics, and artificial intelligence technologies to understand social posts, images, and other unstructured data sources.
- Diagnostic analytics can use IoT sensors to create a digital twin of industrial machinery such as jet engines or gas turbines to enable proactive maintenance or to diagnose failures.
- Descriptive analytics helps answer questions about historic events, such as whether a tennis ball was out of bounds based on triangulated coordinates from synchronized camera images.
- Prescriptive analytics can improve business decision-making by suggestions based on historical and current data.
- Predictive analytics forecasts future outcomes based on historical data analysis. Having the most current data improves confidence in decision-making.
Real-Time Analytics in Retail
Amazon, for example, has boosted its profits by 25% by employing dynamic pricing, which updates product prices every 10 minutes based on competitors’ published prices. Auction websites gather and analyze market prices for goods to suggest prices. Using the right price gets bids and increases sales, helping the auction site and its customers sell more goods.
Real-Time Analytics in Supply Chain Logistics
Food processors and automotive manufacturers optimize space in their factories to maximize production, leaving little space for ingredients or parts. Real-time location systems use RFID tags on trailers to track where they are parked. Using onboard RFID readers to feed updated coordinates of each trailer into a database that analyzes inventory levels in the factory and available dock doors to instruct yard truck drivers to bring specific trailers to an available dock door. These systems ensure production is never stopped due to part or ingredient shortages to maximize output from the facility.
Real-Time Analytics in Transportation
Ride-hailing services match riders to available drivers. For example, the ride-hailing server app collects incoming ride requests at an airport. Drivers are required to wait in a designated parking area which is monitored using a geofence. As drivers enter the geofenced area, they become available for hire. The database-driven server application then matches riders to drivers. This process minimizes wait times for passengers and drivers, avoiding canceled rides.
Real-Time Analytics in Finance
Credit card issuers must approve real-time transactions and protect themselves and their consumers from fraudsters. Every transaction is analyzed for potential fraud by considering the amount, location, and past transactions to see if this is out of the ordinary, often referred to as anomaly detection. Suspicious transactions are declined until verified with the consumer.
In time-critical use cases such as providing stock trading information, every millisecond can make a difference when placing sell or buy orders at market prices. Database latency can become a significant delay, which is why the Elektron trading information platform uses Actian’s columnar vector processing analytic database technology to deliver information to subscribers in under 20 milliseconds.
Real-Time Analytics in Telco
Expandium is an Actian customer that provides network operators with solutions to manage the quality of service. One of their mobile network operators, with over 3 million customers, uses a 12-node Hadoop cluster to provide real-time visibility into network availability. Thanks to the cluster-aware Actian vector processing columnar database technology, the operator can drill down to individual dropped calls. Updates are streamed to the database in micro-batches every 5 seconds.
Providing Real-Time Insurance Quotes
The Automobile Association (AA) provides car insurance quotes that are risk-adjusted and analyzed for potential fraud in online marketplaces. The quote must be delivered in under a second to appear on the first page. Actian’s vector processing columnar database powers the analysis to deliver the fastest quotes with prices that account for multiple risk factors, including the number of vehicle breakdowns the individual has experienced.
How Actian Supports Real-Time Analytics
The Actian Data Platform is an enterprise-proven data integration, data management, and analytics service ideal for real-time applications. You can try it for free for 30 days by visiting our website.