How an AI-powered Feedback Loop Can Drive DevOps
DevOps is a framework that enables organizations to continuously improve upon existing technology assets while ensuring these assets operate as intended. Its quintessential figure eight workflow diagram mimics Agile's focus on short, iterative sprints while emphasizing the separate but equally important role development (Dev) plays in coordination with IT operations (Ops).
Most organizations seek ways to "lubricate" the DevOps workflow, accelerating the value DevOps' Continuous Improvements (CI) give without compromising product integrity. A survey of IT professionals – including DevOps engineers, developers, quality assurance, and site reliability engineers – found that 48% said that manual processes and a lack of tools were their biggest hurdle to CI/Continuous Delivery goals.
Artificial Intelligence (AI) tools can help DevOps organizations achieve their goal of doing more, faster without adding risk. Every "in between" step of DevOps is ripe for automation tools to reduce the need for manual human intervention or review. Simultaneously, AI-powered analytics can give DevOps organizations greater visibility into not just the health of their product, but the true nature of their value stream.
Unify DevOps Data Sets to Enable Better Visibility and Proactive Identification of Improvement Opportunities
Automation is a key selling point of DevOps-focused AI tools, but organizations can achieve foundational benefits before a single process is automated.
Namely: AI-powered analytics can provide organizations with visibility across the entire product ecosystem, enabling easier access to accurate insights. Without a foundational data analytics platform, organizational data is sequestered into their respective silos. This siloing leads to skewed reporting, inevitable bias, and a limited view of factors that may have a greater bearing on the value chain than initially meets the eye.
"Perhaps the biggest impact AI and Machine Learning will have on DevOps is the capability to access and correlate data from disparate sources," writes IT consultant Doug Tedder.
Being able to unify data sets is critical for proper functioning of AI toolsets, which leverage data to train models to think and act autonomously. For instance, a Root Cause Analysis engine (RCA) can rapidly identify the true source of problems and service disruptions – not just their immediate triggers. But without a full view of all organizational data generated from each relevant system of record, a function like RCA could misinterpret events or fail to fully grasp the scope of the associated problem.
Just as importantly, AI-backed analytics can unify the view of the product's entire ecosystem, giving both Dev and Ops a more accurate impression of application functioning at a glance.
"Using AIOps, an engineer can view all the alerts and relevant data produced by the tools in a single place," suggests Tedder, "and the team will have a holistic view of an application's health."
Generate Vital Feedback Through AI-Backed Analytics to Guide Dev Strategy
AI tool sets for DevOps can guide "virtual workers," but they can also make work easier and more productive for humans. For example, a change risk scoring engine can reduce the need for manual review of most changes except for those flagged as high risk. Low-level risks can be automatically remediated while high-level risks can be assessed and mitigated with the appropriate strategic action.
Similarly, a major incident prediction engine can use machine learning to identify which metrics provide the strongest predictive capability for a possible upcoming disruption. Awareness of these risk factors allows operations to monitor them proactively, and, in turn, they can guide Dev priorities to reduce risk factors before code is staged for integration and deployment.
The marriage of AI, analytics, and DevOps shows that the idealized AIOps future has a major role for human operators. DevOps organizations should always prioritize enabling their existing workforce to do more and achieve better results with less latency and fewer limitations.
"Focus on maximising the value of the intelligence your human employees have first," suggests Aaron Hurst of Information Age, "before you start looking to the robots for answers. Once you've done that, look to machine learning and statistics to augment your people, automating away even more of their soul-crushing work in narrow domains such as anomaly detection."
With this guiding principle in mind, DevOps organizations can achieve digital transformation while freeing up humans for innovation, proactive problem solving, and creative strategic thought. Through this approach, the use of AI can help make us more human – in the best ways possible.
Learn more about how AI can empower you to derive more value than ever from your DevOps processes in our recent webinar: "How AIOps helps IT Change and Service Management be more reliable and nimble"