GUEST OPINION by Craig Nielsen, Vice President, APJ at GitLab: Leadership can be a complicated topic. Some management approaches deliver powerful lessons in honest and actionable ways, whilst others are full of corporate jargon, buzzwords and not much else.
My experience running successful technology businesses in APAC has shown me that the best leaders focus on the team, then the problem. This resonates with me, especially now as I face the challenges and opportunities of managing an effective team, with AI already transforming how teams plan, build, deploy, and maintain software.
AI Does Not Replace Strategic Work
Most DevSecOps teams aim to achieve a short time-to-deployment for high-quality software that solves business problems and increases revenue. However, in my experience, too many organisations focus on developer productivity without considering the developer experience. In other words, they’ve got talented developers focused on time-consuming and repetitive tasks, and they perform that work under deadline pressures. While those tasks can be counted, limiting an engineer's productivity measurement to that kind of work can be demoralising.
The good news is that the smart use of AI can remove friction from the software delivery process by taking over less appealing work. This can speed up deployment cycles, improve code security and quality and improve developer morale.
For example, AI can suggest or autocomplete code, perform various tests, or automatically document code functionality in a standard format, all of which would otherwise consume much of the developer’s day.
All of these opportunities equate to a better developer experience. DevSecOps has always been about automation, so why not automate the less appealing tasks?
According to respondents of GitLab’s 2024 Global DevSecOps Report, this shift is underway. They report that AI and machine learning are becoming well-established in software development workflows. A quarter of Australian respondents (25%) spend their time writing new code, with the rest spent on administrative tasks, improving existing code, testing, and mitigating security vulnerabilities. That represents 75% of developers’ day-to-day tasks, where AI can introduce efficiencies.
When AI takes the strain, humans can focus on what they do best: critical thinking and creative innovation. Engineers love tackling challenging projects that test their problem-solving skills. Why not let them concentrate their time on these?
Time For Upskilling
When organisations are intentional with their AI deployments, they can create upskilling opportunities for developers seeking career advancement. Not only does it give them back valuable time to spend on developing new skills, but it can also act as an outstanding coach to them.
For example, AI can impart valuable lessons on optimising code, understanding how it can be better structured and identifying and remediating vulnerabilities before code is deployed. Developers might use AI to learn or to reacquaint themselves with unfamiliar code bases, languages and frameworks.
A 2023 report from global strategy firm McKinsey finds that developers using generative AI-based tools in their work are happier than their peers who don’t have access to these tools. According to the report’s authors, “They attributed this to the tools’ ability to automate grunt work that kept them from more satisfying tasks and to put information at their fingertips faster than a search for solutions across different online platforms.”
These are the developers that every organisation wants to hire, and that’s the kind of developer experience that every engineering leader should aim to deliver. The doers deserve access to the DevSecOps tools they need to get work done and enjoy that work.
In this context, AI seems to be a key ingredient in a DevSecOps solution, critical to an engineering leader’s recipe for success and a powerful way for organisations to attract, engage and retain the best tech talent.
- AI Implementation: Three Planning Considerations
I encourage engineering leaders and development teams to consider three important factors for success. - • Hold your leaders accountable for responsible AI use. I asked my leaders to share how they used our AI features to do their jobs before we asked the teams to change how they work. This benefited our teams in two ways: It required the executive team to engage with the features and experience the challenging parts of incorporating AI into their work, resulting in empathy for change and a shared commitment to ensuring that AI adoption would evolve the way we work.
- • Establish guidelines and workflows to realise the value of AI. It is one thing to ask your teams to improve their productivity with AI, but it's another to be intentional about it. Consider creating a working group to identify best practices and workflows that will change how work gets done. Having teams publish their learnings with before and after comparison data provides insights into how to measure the effectiveness of AI technology.
- • Incentivise learning and sharing. The psychological safety of embracing AI across the organisation can improve when team members share the resources they find useful and their learning journey. This willingness to acknowledge that it is a journey encourages peers to support each other and problem-solve while providing a great opportunity to reward teamwork.
Implementing AI requires careful planning and consideration. You must look at your current business dynamics and the complexity of your current ways of working to determine where AI can most efficiently improve your software development workflows.