As mentioned previously, any move to Shift-Left has to be underpinned by a large amount of structured data linked (or to use the widely accepted term “ETL”, which stands for Extract, Transform and Load) to a long list of complex business rules, which these challenges will put at risk.
Artificial Intelligence however, using a suitably comprehensive and well-designed data layer, can cope with the disparate data created in such an environment, and as it learns from the data will become more effective over time.
The structure of the data layer that facilitates this is called a DataLakehouse and is made up of both structured and unstructured data. Crucially its design combines data with a Metadata and Governance layer, so that in simple terms each data field describes what it contains. This in turn enables the data to be used much more broadly, with less overhead than would otherwise be the case.
The need for a suitable data architecture is therefore crucial to allow an organisation to embrace the potential offered by Artificial Intelligence.
The Value of AI to Service Management
As discussed previously, service management reporting typically references achievements retrospectively, which management then interpret before prioritising and initiating reactive improvement activities dependent upon risk, cost and availability of resources.
What IT management are really looking for though is predictability. After all, service management best practice routinely encourages proactivity over reactivity wherever possible. What we really need is advanced warning of whether our ability to deliver service to our customers is under threat, and if so where does that threat come from, and what can be done to mitigate and remove that threat. Indeed, wherever possible why not allow our well-trained AI capability to implement improvements automatically without human intervention? Removing the underlying cause of incidents before they ever happen has got to be a hugely appealing prospect.
xxx