Why Translytical Data Platforms Are The Way Forward

Jane Fu

The digital economy is all about speed and agility. The modern organization demands analytics at the speed of transactions, but the traditional model of using separate databases, data warehouses, or data lakes to support different workloads is no longer adequate. Having to move data constantly from transactional systems to operational systems and then to analytical systems slows down processing, integration, and timely insights. As a result, reports are stale, data is missing, and insights are already out of date.

Instead, we are seeing the rise of what Forrester1 calls “translytical” data platforms, an emerging technology that combines transactions and analytics on a single platform. These platforms can support several different use cases, including real-time insights, machine learning, streaming analytics, and extreme transactional processing. The key to their success is the ability to perform all of these workloads within a single database. This means, for instance, that businesses can store and process customer data in a single integrated translytical platform, enabling them to upsell and cross-sell new products based on known customer preferences, buying patterns, and previous orders.

Forrester has identified four top translytical database workloads:

  1. Real-time apps, where every second counts. Real-time apps to support operational applications like stock trading, fraud detection, patient health monitoring, and machine analysis have been around for a while. Translytical data platforms can help them work better by providing data 24×7 with low-latency access, as even persisting data can cause unacceptable slowdowns.
  1. Internet-of-Things (IoT) analytics on operational data. When a machine goes down, it can cost a manufacturer millions of dollars every hour — and in some cases, every minute. With IoT sensors, streaming, machine learning, and in-memory technologies, manufacturers can track machines every second to predict likely failures, as well as to decide what parts or resources they might need for repairs if a breakdown does occur.
  1. Connected data appswhere integrated business data is critical. Translytical platforms deliver a real-time, trusted view of critical business data, ensuring that the source of information is accurate to guarantee consistency across the organization. A customer’s address, for example, might be stored on five or more different databases, and a change by one application might not be visible to other app users right away. In this case, storing all customer-critical data in-memory in a translytical database allows all business applications to use it, delivering consistency and integrity.
  1. Continuous learning. As enterprises increasingly rely on machine learning models to make predictions about customers, business processes, and operations, they need to keep these models fresh with new data. Translytical databases support ongoing training, retraining, and monitoring of machine learning models without the need to move data to external machine learning platforms, which is costly and time-consuming.

When combined with a true in-memory database, a translytical data platform can deliver dramatic performance improvements and cost savings across all these use cases and more. IDC2 estimates that an in-memory database can deliver performance improvements of 10x for transactions and up to 100x for analytics. It can also dramatically simplify your data model and streamline your IT landscape, reducing total RDBMS costs (including server, storage, network administration, database administration, and database-related operations) by 40-45%.

As well as providing a next-generation ERP core, a true in-memory platform can double up as a development environment for technologies like automation, machine learning, artificial intelligence, blockchain, and the Internet of Things that we outlined above as they emerge into the mainstream.

But you need to be careful when choosing your in-memory solution. Some vendors merely add in-memory cache to a disk-based database instead of designing it to be 100% in-memory from the ground up. To help inform your decision, take a look at the Forrester Wave: Translytical Data Platforms, Q4 2017, which rates SAP as a leader in this field. “SAP crushes translytical workloads,” the report says. “SAP HANA is a shared-nothing, in-memory data platform, the core of SAP’s translytical platform, which supports many use cases, including real-time applications, analytics, translytical apps, systems of insight, and advanced analytics.”

To learn more, download the full report.

Sources:

  1. The Forrester Wave™: Translytical Data Platforms, Q4 2017
  2. Enabling the Agile Enterprise Through Unified Data, an IDC InfoBrief sponsored by SAP, August 2015.

Jane Fu

About Jane Fu

Jane is on the Global Product and Solution Marketing team for SAP HANA with a focus on advanced analytics capabilities such as spatial, predictive, machine learning, and text analytics.