This article was written by Rich Gibbons of ITAM Review with Peter Rowe, Senior Pre-Sales Consultant, Crayon Australia and Ulrik Roland, Group Vice President, Software & Cloud Analytics at Crayon Group.
Artificial Intelligence is everywhere these days, but it is by no means a new idea.
Talos, the bronze automaton gifted to Europa from Zeus, was an early example and millennia later we had the Mechanical Turk; an 18th century automatic chess playing machine that defeated opponents including Napoleon and Benjamin Franklin before being revealed as just a person in a box. Both are good examples of how AI can capture the human mind and there are plenty of others too – Mary Shelley’s “Frankenstein”, Samuel Butler’s “Darwin among the machines”, Robby the Robot, Rosie from The Jetsons, and Futurama’s Bender.
Firstly, lets define exactly what we mean by Artificial Intelligence (AI) as it’s one of those phrases that everybody thinks they know what it means until you ask them to explain it! Put simply, Artificial intelligence (AI) is the simulation of human intelligence processes by machines. These processes include learning, reasoning, and self-correction. Associated with this is Machine Learning (ML) which is the one of the main methods implemented to achieve AI.
As we look deeper into Artificial Intelligence it may also be useful to consider some of the categories under which it can be implemented, as not all of them are necessarily relevant to IT Asset Management processes, although at least half of them are! The categories we typically see are:
Let’s look at the current challenges that we are all too familiar with in ITAM:
On top of these there are the impacts on scalability, manageability, and complexity of reconciling entitlements with inventory – could it be that AI could offer a solution?
As previously discussed, when we looked at Business Intelligence (BI) and Analytics with reference to SAM tools, all the tools – to a greater or lesser degree – do an adequate job of providing inventory data, reconciling that data to entitlements, and reporting on the outcome of that reconciliation. However, those same tools also produce huge amounts of data and variables and in many cases also store historical data that could be modelled as a ‘trend’ or ‘pattern’ if accessible.
In many cases, SAM tool vendors are now “jumping on the AI bandwagon” and claiming that their tools feature Artificial Intelligence through BI Dashboards, Role and Rule Settings, and predefined “If-Then-Else” workflows and actions. However, in many cases this is no more intelligent than basic pattern definitions and most significantly is not externalised best practice as it is dependent solely on one data set within a single organisation.
A good (albeit non-ITAM) example of everything being “AI” is the story from CES (Consumer Electronics Show) 2020 about an AI pizza machine – that’s just an automated robot. Useful – yes. AI – no.
True “Artificial Intelligence for ITAM” would instead combine that data from many thousands of organisations, making it accessible to all individuals and allow for real predictability based on others’ experiences whilst providing fully customised views.
Specifically, there are several ITAM-specific issues that AI could assist with. The discipline of ITAM is often about ‘firefighting’, and being reactive rather than proactive, as ITAM teams are driven by compelling events coming from outside their jurisdiction and in many cases the ‘theory’ of ITAM being the focal point of the company is not fulfilled.
Therefore, AI offers the opportunity to optimise ITAM / SAM practices in various areas:
AI could be used to interpret a request for software and automate the installation in the way that best utilises the terms and conditions of the publishers’ license or contract terms.
These could be automated in a similar way, with AI not only detecting the breach, but also providing the best resolution, and providing guidance based on license metrics and installation rules.
Again, many SAM and ITAM solution vendors make current claims of AI in terms of procurement automation. However, this is still largely “dumb” orchestration rather than proactive automation that predicts and optimises based on previous trends and patterns, or even external trends and patterns seen previously in other organisations. This Pattern Recognition also has a part to play in wider Asset Management and Maintenance that includes hardware of all types, whereby AI starts linking elasticity around demand with procurement, and even with predictive physical maintenance via monitoring solutions.
Whilst there are concerns that AI further adds to complexity – done properly, it offers a revolutionary opportunity for change. AI offers ITAM better predictability, enables an ITAM practice to move from reactive to proactive and, by linking into other systems, offers the opportunity to truly bring ITAM and SAM into focus at the central point of the organisation.
As with all productivity aids and existing ITAM tools, AI will only improve your ITAM practice if you clearly identify a need, choose the right solution for your specific requirements, implement it correctly, and continue to use and manage the solution. It may be more important than ever to have things correct from the beginning with true AI solutions – if you teach it with poor data and incorrect assumptions, how difficult will it be to get it to “un-learn” all that at a later date? If you install it and then just leave it – who knows what it will be doing in a few months’ time?! Perhaps “AI Systems Manager” will be an ITAM role in the 2020s?