Strategic advice to leverage new technologies

The IT industry struggles to deliver quality, and the majority of effort related to quality improvement is directed toward internal IT processes, rather than the results (i.e., the services) seen by customers.

Along with enterprise architecture (EA) ontology and frameworks, metamodels are one of the most important tools available to the enterprise architect. The potential of a good metamodel compromises a dependence on a range of metamodels provided by vendors, EA modeling tools, or those that form part of an architecture framework. Typical metamodel sources include The Open Group Architecture Framework (TOGAF), ArchiMate, and The Essential Project, while tool vendors include Troux and Sparx Systems. Although there are many commonalities between these diverse sources, there is plenty of scope for improvements and consistency. The accompanying Executive Report examines current best practice to highlight effective metamodeling techniques and suggest ways that EA teams can improve their metamodels and metamodeling.

There are many enterprise architecture (EA) metamodels. Although there has been some attempt to produce a single, definitive metamodel (e.g., TOGAF or ArchiMate), and although EA repository and tool vendors produce detailed metamodels (e.g., Troux or Mega), the topic remains contentious. Is it possible to have a single, all-encompassing metamodel to which everyone can subscribe? How do EA teams reconcile differences between their various metamodels? And how do EA metamodels relate to metamodels in other disciplines, such as ITIL or COBIT? This Executive Report explains why it is impossible to have a single EA model and highlights the techniques currently used to allow for the rationalization of the metamodel maze and improve EA communication through better metamodeling techniques. The report also explains the vital link between the metamodels, views, and viewpoints.

That fresh start may be why BPMSs are a popular corporate enterprise-wide construct to rectify the inadequacies of existing systems and refactor them into the way people really want to do the work. As we discuss in this <i>Executive Update</i>, clouds serve to expedite the process and, at a high level, can appear to deliver the CTO's dreams more quickly. In reality, however, there is a level of complexity to cloud projects that's different from non-cloud projects.

Architecture representations must provide insights that take the enterprise forward, rather than seducing anxious business minds into illusory paths. While we do need to guard against the baser human emotions from hijacking the truth, it is also important to realize that integrity is threatened on a different front: a cognitive one.

The growing ability of big data to seek patterns within immense data streams; perform predictive analysis; and understand language, voice timbre, sentiment, and other attributes thought to require human intervention, raises special concerns.

In this Advisor, we focus on scalability. There are two distinct sorts of scalability: Project scalability -- addressing the ability to perform projects requiring large numbers of engineers; and Organizational scalability -- instituting common development techniques, such as Agile, consistently across an organization. While these are related, they are different. Our fundamental premise is that it is prudent to decide on a prioritization between the two.

Security is a major issue with the Internet of Things (IoT). The more devices you connect, the more you expose the organization to possible attack. And with all the reports of data breaches, system hackings, and cyber-espionage in the news, companies are naturally concerned about security when it comes to building IoT applications and services.