By Manas Sarkar, Vice President and Global Head API Economy, Microservices and NextGen Integration, Infosys
Over the last year, the pandemic has accelerated digital and cloud adoption, as spending on digital technologies continues to grow across enterprises. According to a report by UnivDatos Market Insights, the global digital transformation market is expected to grow at a CAGR of 16.1% from 2021-2026 to reach US$ 3,693.8 billion by 2026.
With cloud becoming mainstream, organizations are adopting cloud apps to accelerate digital transformation of their processes and deliver superior experiences. In turn, this has necessitated the integration of data, process, and experience across these cloud apps through cloud-first integration and automation.
The concept of process integration is not necessarily new since it also existed in the traditional on-premises setup. However, the adoption of cloud apps has added considerable complexity to process integration since it presents several challenges around:
• Multiple Platforms in Anything as a Service (XaaS) economy
In the traditional on-premises setup, the business process transformation process typically involved the implementation of an ERP platform, which accounted for about 70 – 80% of all business processes. This does not hold true in the case of cloud apps economy, which are typically very focused on niche business capabilities. Each cloud app generally accounts for only 30 – 40% of end-to-end business processes. Therefore, implementing any process transformation at full length is likely to involve three to four different cloud apps platforms. These then need to be integrated in the context of end-to-end business processes, which could run across multiple cloud platforms.
• Need for Customization
Specialized out-of-the-box cloud apps are likely to address up to about 60 – 70% of any business requirement. Beyond that, these need to be customized to meet unique business requirements. Unlike traditional ERP implementation, customization of the core cloud apps is extremely complex since they are provided via a service delivery model. Therefore, software vendors tend to keep the customization of the cloud apps separate from the implementation process. For example, SAP generally recommends its Business Technology Platform (BTP) for implementing customization for S4 implementations.
• Disparate Data
Every business process requires data – whether it is master data or data from related process areas or market data. Unlike in a traditional on-premises set-up where the data resides locally, cloud data can reside anywhere. It could either be housed within a data lake implemented in one of the hyper-scalers or it could be in another cloud app or even on premise.
• Secure Access
With process fragments and data fragments lying across different clouds, ensuring security while providing seamless access to cloud apps and data across the end-to-end business process becomes a major challenge.
Implementing Seamless Process Integration
Addressing some of the integration challenges posed by cloud apps requires a considerable focus on process automation and integration capability. Let us take the example of supply chain integration. The end-to-end value chain of a supply chain is likely to involve multiple cloud apps such as Supply Chain Planning (SCP), Order Management Systems (OMS), Transport Management Systems (TMS), Warehouse Management Systems (WMS), Merchandizing Systems (Merc) and B2B interactions.
In addition to this, the supply chain process needs to integrate seamlessly with upstream process such as Configure, price quote (CPQ) and eCommerce. Also, functions such as SCP, WMS, OMS, TMS, and Merc are generally implemented as separate cloud apps. At the same time, there is also a tremendous push to modernize B2B exchanges. To add to this, the supply chain process needs a lot of master data related to customer, partner, and products, which are generally implemented in different data lakes. Therefore, any supply chain process integration needs to address various aspects such as:
1. The ability to provide ‘digital bonding’ or the integration of desperate systems seamlessly in a stateless / stateful manner using different API architectural styles or techniques such as representational state transfer (REST), Simple object access protocol (SOAP), GraphQL, Electronic Data Interchange (EDI), Messaging and more
2. Seamless integration across B2B, web, mobile, and customer service channels
3. Process automation for seamless process flow
4. The capability to support process and data extensions for the supply chain process, which could mean the ability to build microservices to extend service or process apps to handle exceptions
5. Extensive visibility for end-to-end processes with the ability to proactively identify process breaks
6. The ability to get micro feedback across the process chain to identify opportunities for innovation.
To seamlessly cater to a wide variety of requirements such as those listed in the example above, there needs to be a cloud-first integration and automation approach that is equipped to handle the level of complexity that integration of cloud apps presents. As the adoption of cloud apps continues to grow, evolving the process integration process to reflect the new reality and challenges is crucial.