Check in the Box: AIOps
Check in the Box: AIOps

Check in the Box: AIOps

It’s all the buzz right now and we have even seen tenders asking for system compliance to AIOps, like it’s a check in the box kind of thing. I cringe at the thought! One of the challenges that we list when talking to clients about AIOps is the lack of strategy. It’s not a “check-in-the-box” type of feature, although sometimes it is marketed as such. 

We need to think about the practical implementation of AIOps, as well as future, pie in the sky visions of what it can become. The far-fetched wish list takes more planning, integration and implementation to achieve, with smaller gains along the way. The goal is to achieve systems automation to replace employees that are locked into performing mundane tasks. Or is it to save costs of the staff doing those tasks by implementing a more affordable system? 

Back to strategy: we love having these conversations with our clients because it requires a bit of thinking. What are the actual tasks that you need automated? Where? There are multiple places in organisations that may be eligible. Do those tasks require intelligence based on specific inputs? Are these inputs available as data feeds from a system to ingest? How fast is the current process? Who or what are the current bottle necks? Are there volume issues that delay processes? What is the process that that we’re trying to improve on? 

I could go on – but you get the picture. 

Many times, process automation is not something that should be sought after within a specific system but rather across multiple systems. If you think about the number of systems you access to get your work done, it’s similar. Yes, certain systems could do with better features, but imagine if they were all integrated and you could build a process model across all the systems to make your work lighter, faster or more efficient. 

When companies ask for “AIOps” as a checkbox feature, most companies can tick that box, as they have elements of automation built into their system – but until we really start thinking about automation elements with our businesses and specifically asking for them, we’re not going to comply to well thought-out strategies. 

Other points to consider when formulating your AIOps strategy:

  • Machine Learning: Does it have a place and if so, where? This avenue can open up into an entire new strategy to complement AIOps; not to be underestimated. 
  • Human capital training investment: Upskill your staff and don’t underestimate their investment required to make AIOps work effectively. 
  • Data collection: Remember that we need data to ingest for decision-making, so where is it coming from, what format is it in etc.
  • Data standards, enrichment, staging and storage: Multiple data feeds and the normalisation of data may need a new storage area or integration platform for ease-of-use and enrichment. 
  • APIs, integration, and collaboration: Remember to review all systems APIs and/or or integration capabilities. APIs need to be mature and easily accessible.
  • Architecture: Remember that when systems are integrated or when data is extracted, you should have a flexible, yet solid design that is well-documented and maintained. 
  • Finally, consider the costs and return on investment. It’s no use investing huge amounts of money for minimal return(s). Target significant changes that are worth the investment. 

If you explore the market for an automation partner, please don’t ask for a compliance tick. Share your strategy and allow vendors to understand your short- and long-term milestones. 

Emile Biagio

CTO 

Last modified on Friday, 30 September 2022 05:45
(0 votes)

Leave a comment

close