Business has become faster. Not just because of connectivity and the fact that the world’s financial systems are more tightly interconnected than ever before. It’s also thanks to the IT department that works to orchestrate the data and applications that we use every day.
The fact that we’re now all running on data-driven, software-centric operational backbones brings massive opportunities for efficiency and new economies of scale. But this new reality also comes with some important caveats.
As every business in every industry is being disrupted by software, organizations are trying to figure out how to navigate and embrace these new digitally transformative forces. What this creates is an important inflection point in terms of how we architect, build, test, deploy and manage our software.
Point of inflection
The inflection point is easy to spot. Software engineers — developers, architects, testing professionals, AI specialists, all of them — have long held an incongruous relationship with the operations department professionals — systems administrators, database administrators, configuration management gurus, all of them — who support the applications and services created.
Developers want to speed up innovation cycles and deliver software faster, continuously. But operations want to use tried and tested means to manage that process, which can appear cumbersome, laborious and sluggish to the developers. The operations people know that rapid changes to business systems without proper safeguards can impact business operations, but developers can be impetuous in their rush to create spontaneous creative new tools that they hope will change the world.
Add these issues to the vast number of tools used by developers and IT, and we have a maelstrom of factors that are all applying forces in different directions. The variety of form and function here makes it difficult for leaders to get a holistic view of how each team is performing. This lack of visibility slows the behavioral and cultural change needed to succeed.
Thankfully, we have over the course of this decade refined an approach that brings the combined efforts of developers and operations more closely together. We have called it, very logically, DevOps. Forrester Research suggests that three in four businesses will adopt DevOps methodologies in the next year in order to strike a balance between moving fast and ensuring the business runs smoothly.
As we turn the decade then, DevOps is with us and we can see that different organisations are implementing it at differing degrees of success. So what happens next?
Diminishing DevOps dilemmas
What matters now is how we use DevOps to streamline software development. Being able to make it ‘fit’ the business — and I mean precision engineering level fit — entails being able to scale it across the entire organization’s technology estate. This is one of the key DevOps dilemmas that the majority of companies finds most challenging to achieve.
With many ‘ops’ functions being performed manually, the implementation of enterprise processes in audit, compliance, security and change management needs to be elevated to a higher plane.
But it’s not just the burden of manual process — older firms are weighed down with a plethora of systems that have no interconnectivity or integration. There is duplication of data and a lack of care in terms of keeping data pipelines current, which means that new automation methods don’t necessarily work without massive investment and re-engineering. This introduces huge risk at a time when customers are fickle and can leave in a heartbeat.
Organizations looking to scale the application of DevOps need to take a platform-level view of their deployments. A more comprehensive and connected approach to DevOps can provide the required level of visibility across the entire product development toolchain. Focusing on a single part of the DevOps process is good, but doing so within a broader dashboard view of all technologies in play is essential if organizations are going to be able to plan, develop, test, deploy and then further develop.
Financial services under pressure
We see massive pressures put upon established firms in the financial services industry due to huge regulatory changes happening across the sector. These companies need to bring legacy systems forward into digital and do it at a massive scale in controlled secure environments.
At the same time, the established players are being disrupted by fintech startups with no legacy systems to overhaul. These smaller startups are naturally less monolithic and are often brilliant at delivering a single product or service. This means they move so much faster and in tune with rapidly changing customer needs. The startups — in many cases DevOps natives from the get-go — are providing further proof for the established bastions of financial services that they need to adopt DevOps in order to go forward.
DevOps is here, now onto DevAIOps
Positive, productive and pragmatic DevOps is a working environment where business leaders can easily view common metrics across teams and drive best practices, regardless of the tools being used. This can mean shorter development cycles, increased deployment frequency and more dependable, high-quality releases.
Truly effective DevOps allows IT departments to approve code faster by automating the process based on pre-set rules and policies determined by change managers, in turn allowing them to focus attention on higher risk submissions. Developers should be able to see their code in production in minutes rather than days or weeks, and build code and submit it for approval using their favourite tools.
As we embrace DevOps, we can start applying a growing number of autonomous tools that provide self-checking and self-healing mechanisms — adding AI to enable DevAIOPs. Code gets fixed, updated, augmented and deployed without human intervention. This frees up operators on both sides of the DevOps divide to focus on more complex issues that are truly business impacting.
Automation is a key part of our DevOps future, but it won’t replace the humans or signal the rise of the robots. All enterprises must gear up for DevOps to help the business move faster. At the same time, DevAIOps is on the way and we need to ‘smarten up’ for it to drive business to the next level.