Richard Whitehead discusses automation, the steps organizations can take to implement autonomous processes, and the main barriers facing enterprises, the role full and partial automation will play in the future, and how many IT managers will need to tackle automation. Learn some important skills and how you can do it. Use AIOps to improve availability and support, further detailed below.
ADM: Do you envision a future where everything is fully automated? Why or not?
Whitehead: I don’t Full automation is what everyone wants, but I would say nobody wants it. This is because organizations using full automation can turn the slightest mistake into a catastrophe without humans having any idea what is going on. This is terrifying. So given the complexity and fragility of IT systems, many teams want to intervene or use it as a last resort. In an SRE environment, an engineer may limit her push of automated code to protect her budget from errors. Partial automation is a more achievable goal than full automation and a top priority for technology leaders. In AI, we call this the “man in the middle”.
ADM: We often hear fears that automation will take jobs. Is it a legitimate concern? why not?
Whitehead: Automating human work is not a problem we should worry about right now. Businesses need to realize that automation supports IT teams, not replaces IT jobs. Automation takes repetitive tasks off the plate of SRE teams and DevOps practitioners, freeing up time to focus on more valuable tasks. For example, automated solutions such as artificial intelligence for IT operations (AIOps) scan large amounts of data and detect incidents that may affect services before they affect end-users. can. Without AI, it’s like finding a needle in a haystack, impossible even for the best resourced teams. But even this technology cannot replace human teams. Rather, the platform notifies teams when there are issues in the IT ecosystem that need human ingenuity and creativity to fix.
A recent Dell study found that organizations are digitally transforming, but are slow to adopt automation. As IT environments become more complex, the need for automation becomes more and more important, and hesitant businesses must embrace automation to keep up with digital transformation.
ADM: What is stopping organizations from implementing autonomous processes?
Whitehead: The reason organizations are reluctant to implement automation is simple. It’s because people are nervous. Change is hard and people crave human oversight. Even in a fully automated world, people still like notifications and buttons to press.
- Trust: Teams need to prove that automation works before they can trust the system to do the right thing.
- Budget constraints: Implementing automation can be costly and the process has to be done frequently to justify the cost. I believe in the “Rule of 3”. If it happens 3 or more times, automate it.
- Lack of in-house expertise: Writing automation code can take time and effort to set up. Implementing automated processes can be a struggle if your company does not have an IT team skilled in automation.
- Security: It is imperative to ensure that automation cannot be abused by malicious actors. For example, expensive and time-consuming automation can be a boon for those who want to recreate the effects of a DDoS attack. Think like Hans Gruber in Die Hard. He has exploited his known FBI correspondence to achieve his true goals.
ADM: What is partial automation and how do you determine if it’s the best strategy for your organization?
Whitehead: Partial automation is setting up automation with a “man-in-the-middle” approach. As I mentioned earlier, even DevOps practitioners still like to push the button. Partial automation allows human operators to work with technology to produce results, customize algorithms, and train technology to repeat desired actions. Building a partial automation strategy can help your organization step into automation and give your team more confidence in automation tools.
ADM: What are the ideal skills an IT manager should have when working with automation?
Whitehead: Organizations need team members who can write code and understand the backend of automated processes. But when it comes to management, mindset is everything.
Successful automation requires leaders to trust their tools and maintain a growth mindset. This is especially important for the leader of his SRE and DevOps team, which is small and can only be scaled with the help of automated tools. Scalability is likely why his 36% of organizations with higher automation maturity expect revenue growth of 15% or more for him, according to a Dell report. Similarly, only 10% of organizations with lower automation maturity expect the same growth.
ADM: What small steps can organizations take to implement automation in their workflows?
Whitehead: Most importantly, focus on getting your IT operations team on board. Where resistance to change breeds reluctance to automate, use partial automation to start building trust between your IT team and technology. Ask a team of humans to train the technology in the same way you would a pet, guiding the tool to its ideal outcome through positive and negative feedback. Use semi-supervised automation that allows teams to customize ML algorithms. And finally, consider employing tools that allow users to peek behind the curtain and see how technology works.
ADM: Automation often leads to higher availability. How does AIOps support automation and availability?
Whitehead: Two key features of AIOps support automation and availability. ML and Natural Language Processing (NLP). AIOps solutions identify data anomalies that indicate potential service-impacting incidents and alert SRE and DevOps teams that there are issues to address. But AIOps is much more than that. ML capabilities learn disruptive patterns and prevent them from happening again, reducing incidents and outages and their associated unplanned work.
With NLP capabilities, AIOps understands the nuances and flexibility of language. Previously, only humans could tell 10 North First Street from being the same as 10 N. 1st St, creating a very complex infrastructure of rules. But with NLP, code can go beyond raw matching and see words, understand English, and make correct correlations.
These features mean AIOps technology is flexible and robust in the face of change. Automation tools with this kind of staying power are much easier to justify economically than old and fragile automation tools.
About Richard Whitehead
As Moogsoft’s Chief Evangelist, Richard brings a keen sense of what it takes to build innovative solutions. Former CTO and VP of Technology, Richard brought new technologies to market and was responsible for strategy, partnerships and product research. Richard served on Splunk’s technical advisory board throughout the Series A, providing product and market guidance. He has served as an ambassador for his DevOps Institute and most recently he co-chaired the ONUG Monitoring & Observability Working Group. Richard holds three patents and JavaScript is considered dangerous.
