Organisations have pushed the button on their digital transformation journey, and now understand that they need to be mindful of how they integrate and manage enterprise data that is distributed, still easily accessible, trusted, and governed. In fact, Gartner predicts a 6.5% growth rate across the board for the IT sector in Australia, with the biggest growth expected to come from the IT services sector.

Coupled to that growth, organisations are placing emphasis on their back-end systems in order to get their customer-facing assets working better. In many cases this means migrating to more efficient data solutions, removing siloes, ‘cleansing’ data and generally making better use of the data assets the company has.

This has prompted the advent of modern data integration styles like data virtualisation, as industries of all kinds and sizes look to accelerate change and leverage data more effectively.

Expect to see the following five trends make their mark in 2022:

1. Data fabric becomes the foundation for the distributed enterprise. As digital businesses and online sales channels proliferate and remote work becomes the norm, it creates a complex and diverse ecosystem of devices, applications, and data infrastructure. In particular, data infrastructure can span on-premises, single cloud, multi-cloud, hybrid-cloud, or a combination of these, spread across regional boundaries with no single solution to knit all this data together.

In 2022, organisations will create a data fabric to drive enterprise-wide data and analytics and to automate many of the data integration, preparation, exploration tasks. Data Fabric unifies the data assets distributed across disparate location, format, and latency using logical, physical, or hybrid approaches. By enabling organisations to choose their preferred approach, these data fabrics will reduce time-to-delivery and make it a preferred data management approach in the coming year.

In fact, according to a recent TEI study by Forrester, “Data fabric technology takes data virtualisation a step further by automating data management functions using artificial intelligence/machine learning and providing additional semantic capabilities through data catalogue, data preparation, and data modelling.”

2. Decision intelligence makes inroads for enterprise-wide decision support. Organisations have been acquiring vast amounts of data and need to leverage that information to drive business outcomes. Decision intelligence is making inroads across enterprises, as regular dashboards and BI platforms are augmented with AI/ML-driven decision support systems. Decision intelligence is the combination of regular BI dashboards enhanced with AI/ML, whereby enterprises can make predictions of outcomes for certain sets of actions and recommend one action over the other, thus helping decision support systems.

In 2002, decision intelligence has the potential to make assessments better and faster, given machine generated decisions can be processed at speeds that humans simply cannot. The caveat - machines still lack consciousness and do not understand the implications of the decision outcome. Look for organisations to incorporate decision intelligence into their BI stack to continuously measure the outcome to avoid unintended consequences by tweaking the decision parameters accordingly.

3. Data mesh architectures become more enticing. As organisations grow in size and complexity, central data teams are forced to deal with a wide array of functional units and associated data consumers. This makes it difficult to understand the data requirements for all cross functional teams and offer the right set of data products to their consumers. Data mesh is a new decentralised data architecture approach for data analytics that aims to remove bottlenecks and take data decisions closer to those who understand the data.

In 2022 and beyond, larger organisations with distributed data environments will implement a data mesh architecture. As different functional units or domains within larger organisations have a better understanding of how their data should be used, letting the domains define and implement their own data infrastructure results in fewer iterations until business needs are met and are of high quality. This also removes the bottleneck of the centralised infrastructure and gives domains autonomy to use the best tools for their particular situations. Data mesh will create a unified infrastructure enabling domains to create and share data products while enforcing standards for interoperability, quality, governance, and security.

4. Organisations embrace composable data and analytics to empower data consumers. Monolithic architectures are already a thing of the past but we can expect even smaller footprints. As global companies deal with distributed data across regional, cloud and data centre boundaries, consolidating that data in one central location is practically impossible. That’s where composable data architecture, whereby organisations can pick and choose certain tools to build parts of or the entirely of their data infrastructure, becomes paramount and brings agility to data infrastructure. One good example of a composable architecture is a data fabric, which can be created using a data catalogue tool, a semantic tool, a data integration tool and a metadata tool put together.

Data management infrastructure is extremely diverse and usually every organisation uses multiple systems or modules that together constitute their data management environment. Being able to build a low-code, no-code data infrastructure provides flexibility and user friendliness, as it empowers business users to put together their desired data management stack and makes them less dependent on IT.

In 2022, expect organisations to accelerate building composable data and analytics environments, whereby they can avoid vendor lock-in and attain more flexibility as they put together a data infrastructure stack that meets their needs.

5. Small and wide data analytics begin to catch on. AI/ML is transforming the way organisations operate, but to be successful, it is also dependent on historical data analytics, aka big data analytics. While big data analytics is here to stay, in many cases this old historical data continues to lose its value.

In 2022, organisations will leverage small data analytics to create hyper-personalised experiences for their individual customers to understand customer sentiment around a specific product or service within a short time window. While wide data analytics, which entails combining structured, unstructured and semi-structured data from various data sources for analytical purposes, is a comparatively new concept and yet to find widespread adoption - given the pace at which organisations are making use of geospatial data, machine generated data, social media data and various other data types - expect to see small and wide data analytics gaining better traction across organisations as we enter into the New Year.