Four requirements of the modern enterprise data infrastructure

In the current digital economy, applications require a myriad of data sources from different devices, servers and machines. This data is necessary to build innovative solutions and create competitive operational efficiencies. Unfortunately, this onslaught of unstructured and structured data comes from multiple locations and is forcing many organisations to create complex “hodge podge” data infrastructures to support their business needs.  

From small credit unions to large automated warehouse systems, organisations are increasingly looking for ways to simplify their data architecture and quicken time to value from data. And in today’s environment, a modern data platform needs to be consolidated, convenient, and reliable. There are four elements that organisations must adhere to in order to create a modern data infrastructure: the database, data analytics, data and application integration and decision workflow. While these traditional data technologies aren’t new, it’s how they work as a single combined entity that will ensure success in today’s data-driven world. 

The database

The database remains the foundation of a data platform. However, over the course of the last decade, the requirements for supporting multiple data types and multiple workloads in the same business process while interoperating with existing legacy business processes have changed. The data and compute landscape that supports those processes has shifted and become extremely complex. As a result, the modern data platform should aim to take advantage of all possible sources of value while removing complexity.

Mid-size banks and credit unions, for example, are facing competitive threats from online and mobile banking models and are taking a different approach to customer engagement. They are focusing on customer experience and building relationships through live interactions powered by purpose-built CRM systems. Thus, they have the need to analyse transactions, develop offers tailored to their customer base, and empower branches with smart outreach programs – all with limited IT budgets and technical staff.

Implementing a single, unified data platform that supports both transactional and analytical business processes and workloads eliminates the need for complex data workflows and removes latency introduced when data sets between workload-specific systems are copied. It becomes difficult to support tactical, time-sensitive business decisions if data is segregated into transactional and analytical databases. Thus, having a consolidated data platform is not only convenient and cost effective, but also critical for applications where bank branch managers are interacting with customers in their office, at the counter or in the drive-through.  

Data analytics

Data analytics also plays a significant role in the modern data infrastructure. For most organisations, some level of analytics is expected in any new application to help users make quick and accurate decisions. Today’s analytics in particular need to be able to run concurrently when a transaction is processed, and not be restricted to a specific type of data (for example, structured, unstructured, streaming, etc.). The ability to analyse data in real-time and turn the new information into action is also vital. Real-time analytics enables organisations to gain insights to help shape important business and financial decisions that drive business growth, detect fraud, and assess risks in real-time.

In the case of the bank branch manager, how do they know the risk profile or profitability of a long-time customer asking for a home equity loan? Without embedded analytics that can parse through credit scores and risk analysis feeds with online application data, the branch manager cannot make a timely, informed decision. 

When it comes to ensuring effective data workflow with data analytics, it helps to erase the need to move data from a transactional system to an analytical system. Administering a unified data platform removes this step and helps ensure that you are making an informed decision based on the latest data, versus making a decision based on stale, outdated information. For many organisations, especially those that dabble in trade verification, billing, sensor management and fraud detection, basing important actions off of outdated information can be crippling. 

Data and application integration

As data continues to inform increasing amounts of corporate strategy, organisations must be capable of effectively incorporating new data sources without missing a beat. Organisations realise that to be effective they must integrate data into their decisions and actions from day one, and many are utilising APIs to improve data integration and flow across the enterprise through existing and new composite applications. Looking at data and application integration today, flexibility is critical as new data sources and new cloud-based SaaS applications are being introduced to businesses at a mind-numbing pace. Data interoperability, or supporting the ability to add, augment, and swap out data sources without disrupting live production applications and analytics, is crucial for the agile enterprise. With such an enormous volume of data being generated by an array of IoT devices and new applications, the integration, management and connectivity involved is often overlooked.

Following our example of the simple bank transaction decision, new ways to detect fraud and calculate risk are made available to retail banks via online SaaS-based applications. Falling behind on these new innovations can expose banks to new and unnecessary risks.

Decision workflow

To understand the impact a decision can have on dependent systems, current complex business problems require more real-time feedback loops and the ability to support iterative learning. In today’s world where machine learning technologies are booming, organisations need to put new insights into immediate action to receive impactful business results. Organisations should also consider reducing latency between when information is available and when it is used to make a decision, as the results are then amplified. For example, when bad news travels slowly, the impact is worse than if you have the bad news break immediately so you can course correct. For our bank customer inquiring about a loan, and the bank making a decision between a profitable versus risky transaction, decision workflow involves making sure both parties have all the information they need to take the next best action – one that cements the trusted relationship.

To ensure an effective data workflow, more integration points with real time feedback loops inserted into the decision making process is invaluable. A data platform that can support a complex system that enables analytics and informed insights at the time a decision is made, benefits businesses tremendously.

As the vast amount of data, devices and applications continue to grow every day, organisations will need to be wary of succumbing to old habits of complex, inefficient data structures and outdated integration techniques. Looking at these four elements to the modern data infrastructure, it is clear that simplicity, convenience and speed is key. Each and every one of these components are valuable to the modern data infrastructure, and will help ensure success in today’s data-driven world.

Julie Lockner, InterSystems
Image source: Shutterstock/everything possible