Video Interviews Category - Page 7 of 7

How AI Is Transforming the Future of Software Development with Wing To

Digital.ai CTO Wing To delves into how the way code is written, deployed and managed will fundamentally change in the artificial intelligence (AI) era.

Cutting through the extremes, AI won’t erase entry-level dev jobs, nor will it fully replace senior developers. Instead, it’s changing what both groups focus on. Entry-level devs will need to learn how to effectively prompt, evaluate and refine AI-generated code, while senior devs may shift toward mentoring, reviewing AI-assisted output, and ensuring software quality. The fundamental knowledge of how systems work still matters—AI doesn’t remove the need for strong foundational skills.

But AI does change the coding landscape. Developers are writing less and reading more. Reviewing AI-generated code—often verbose, not elegant—has become part of the job. And while some fear that the code is harder to debug, it may just be different: more mechanical, less idiosyncratic.

From a DevOps perspective, AI introduces new challenges and opportunities. DevOps engineers are facing a surge in code volume and speed—often without a corresponding bump in quality. Automation, golden paths and better processes are essential to handle the influx of AI-assisted code. AI can also help streamline repetitive work: writing scripts, maintaining configurations, generating tests or spotting compliance risks.

Eventually, DevOps teams may evolve into ecosystems of human engineers and collaborative AI agents. That vision’s still emerging, but orchestration between tools and agents is the next hurdle. The goal? To let developers and engineers focus more on creative, value-adding tasks—and less on toil.

Still, AI must be used thoughtfully. It’s not a magic fix. It’s a power tool—one that needs skill, discipline and a good understanding of when (and how) to use it.

June 10, 2025

GenAI in Software Development with Nathen Harvey

Nathen Harvey, DORA Lead at Google Cloud, delves into how generative AI is rapidly reshaping the software development landscape, presenting both exciting opportunities and complex challenges.

Harvey reflects on his own journey, from early web operations and his time at Chef to leading the DORA program after its acquisition by Google. DORA (DevOps Research and Assessment), now in its 10th year, continues to be a definitive source for understanding how technical teams improve performance and value delivery. The program’s annual reports, especially its latest focused on generative AI, offer research-backed insights that are widely used across the industry.

The conversation dives deep into this year’s findings, starting with the rapid and widespread adoption of AI. Nearly 90% of organizations surveyed are prioritizing AI from a top-down perspective, while over 76% of practitioners report already using AI tools daily. This simultaneous push from leadership and uptake by developers indicates that AI isn’t a passing trend—it’s a permanent shift.

But the data reveals nuance. While AI tools are helping improve areas like documentation, productivity and developer satisfaction, there’s also a measurable decrease in software delivery stability and throughput. Harvey suggests that this may be due to larger and more complex change sets generated by AI tools, which clash with workflows optimized for smaller, more iterative updates. In essence, we’re accelerating one part of the pipeline while other parts can’t yet keep up.

Harvey emphasizes the importance of interpreting DORA findings in context—using them as hypotheses to test within each unique organization. He also raises critical questions about workforce development, particularly how AI might reduce opportunities for junior developers to gain real-world experience and grow into senior roles. The future, he notes, will require balance: leveraging AI as an amplifier while still investing in people and process.

The takeaway? AI is transforming DevOps, but not without growing pains. To navigate this shift, organizations need to blend data-driven research with community collaboration, continuous experimentation, and a commitment to developing talent at every level.

April 23, 2025

Feature Management’s Role in DevOps with Trevor Stuart

Trevor Stuart, general manager and senior vice president of Harness, explains how feature management is becoming a critical extension of DevOps workflows a year after the acquisition of Split Software.

Feature management is fast becoming a critical pillar in modern DevOps workflows, especially as organizations seek to keep up with the rising volume and velocity of software development. Stuart discusses how the evolution of feature flags has reshaped deployment practices and the expectations of software delivery teams.

The core value of feature flags lies in their ability to enable safer, faster releases. They give teams the flexibility to toggle functionality without redeploying code, making it possible to mitigate risk during production rollouts. Initially used sparingly and often managed via manual config files, feature flags are now a mainstream method for progressive delivery. They help organizations validate features with specific user segments, reduce blast radius when issues arise, and support experimentation in real-world conditions.

Stuart highlights that the growing use of AI-generated code, modular microservices, and increasing product personalization have only amplified the need for robust feature flag systems. These tools now serve not just as toggles, but as real-time control layers for dynamic configuration, adaptive experimentation, and telemetry-based feedback. Companies are increasingly using them to fine-tune AI models, personalize customer experiences, and refine software quality using real-time performance data.

Despite broad awareness, many teams still rely on homegrown or brittle solutions, leading to maintenance issues and outages. The challenge is often organizational: determining whether to build, buy, or integrate these capabilities into broader CI/CD pipelines. Stuart argues that deeply integrating feature flagging into delivery workflows—particularly as part of a unified platform—is essential for scale and consistency.

As software complexity grows, so does the need for intelligent automation. Stuart envisions pipelines that not only deploy features but also observe impact, run experiments, adapt based on results, and even feed signals back into models. It’s a shift from manual gates to data-driven automation, marking a pivotal transition in how DevOps teams manage risk, velocity, and innovation.

April 23, 2025

Building Complex Software at Scale with Tameem Hourani

RapDev founder Tameem Hourani dives into what it really takes for software engineering teams to develop and deploy complex software at scale.

Hourani talks about the challenges of maintaining a blameless culture in DevOps, even as software complexity continues to grow. The discussion explores how teams can strike a balance between rapid innovation and system resilience while avoiding a culture of finger-pointing when failures occur.

He also emphasizes that failures are inevitable and should be treated as opportunities to strengthen systems rather than assign blame. He discusses strategies for building resilient architectures, such as feature toggles, blue-green deployments, and A/B testing, which allow teams to roll back changes quickly and minimize downtime. The key, he argues, is to create systems that can adapt and recover efficiently rather than trying to prevent every possible failure.

The discussion then shifts to the role of observability and automation in modern DevOps workflows. With AI generating more code than ever, organizations must ensure robust testing and monitoring practices remain in place. Hourani highlights the growing use of AI in both detecting and fixing issues, potentially enabling systems to self-correct without human intervention.

They also examine the broader evolution of DevOps, including the shift toward developers managing more aspects of infrastructure. As platform engineering gains traction, organizations are working to centralize shared services while enabling developers to have greater control over their environments. Security is also becoming a more integrated part of DevOps, with policies and guardrails moving into GitOps workflows rather than being managed separately by security teams.

Hourani concludes with insights on common pitfalls in DevOps, including misconceptions about infrastructure-as-code solving all scalability issues and the ongoing challenge of balancing speed with stability. Overall, there is a need for cultural shifts, automation, and smarter engineering practices to keep pace with the demands of modern software development.

March 6, 2025

2025 Predictions for DevOps and Application Development

Hi everyone. I’m Mitch Ashley, VP and practice lead for DevOps and application development with the Futurum Group. These are my predictions for 2025.

First, 2025 is go-time for AI in production. A lot of AI projects have been stuck in the pilot and prototype phases. Matter of fact, some estimates say that only a third of those have made it into production so far. This is the year where AI has to get real and start demonstrating value. That impacts how we develop traditional software as well. We will see vendors introduce capabilities that help us start to blend and coordinate workflows and pipelines across DevOps for both traditional software as well as AI projects, which tend to have some differences in how the workflows are performed. So look for those features maybe later in 2025.

Prediction number two, built-in AI, rather than bolt-on AI across the software development lifecycle. Much of the AI we use, particularly generative AI today, are things that are bolt-on. They’re chatbots, they’re natural language interface to it. They may be an IDE plugin that helps us generate code or get access to a code base. Increasingly, we’ll see more AI that is just a part of the products and the tools and the workflows, the tool chains that we use for creating software rather than an adjunct or separate thing. I think we’ll see a lot more productivity as more AI is really integrated into how we work rather than a feature of a product or a separate product.

Kubernetes dominance, that’s the next prediction. We tend to think of Kubernetes as the cloud-native container orchestration software. Of course, that’s what it is. We use it with microservices, containerized apps, all kinds of workloads. But that’s the point, is that Kubernetes has really expanded its use across virtually any kind of product and service that we may use. We may be provided by a third party, a service in the Cloud. It could be even on hardware, integrated hardware and software to use for storage solutions or in ed solutions. Kubernetes truly is everywhere and it has become the workload system for any kind of application or service. With that comes though, we need to up our skills in Kubernetes operational capabilities. It is complex. It’s getting simpler. We’re offloading more of that to third parties. But we’ll also see AI help us here with operationalizing Kubernetes, reducing some of the complexity and improving our overall ability to run it effectively.

Well, thank you for joining me for these predictions around DevOps and application development. I hope you’ll check out the full ebook of all the predictions by The Futurum Group analysts. We cover a number of areas, whether it be enterprise applications, AI, of course, infrastructure, security, you name it. We’ve got some great analysts and I hope you check out those predictions at Futurumgroup.com. Thank you.

February 18, 2025

Open Source Software Security Concerns with Spike Curtis

Spike Curtis, principal engineer for Coder Technologies, dives into why open source software security concerns are valid, and why the only viable option is to invest more in securing software supply chains to mitigate potential threats. While open source benefits from community oversight, organizations must vet the projects they rely on to avoid security gaps.

Curtis also takes a look at AI’s role in software security. While AI has potential for detecting vulnerabilities, current tools aren’t yet reliable for systematic security checks. He warns that AI-generated code can introduce more security flaws if developers prioritize speed over careful review. Addressing concerns about enterprise contributions, he notes that while some companies actively support open-source projects, many simply consume without giving back. Organizations depending on underfunded tools should consider dedicating resources to their maintenance rather than assuming the community will sustain them.

February 6, 2025

Anjali Arora on DevOps Teams Embracing Platform Engineering

Anjali Arora, chief technology officer at Perforce Software, discusses the inefficiencies in DevOps infrastructure and the role of platform engineering in addressing them. She notes the shift from traditional IT-managed infrastructure to DevOps-driven provisioning, and how it’s led to a fragmented and costly environment. To counter this, organizations are turning to platform engineering to strike a balance between speed and governance, ensuring that DevOps teams maintain agility while optimizing resources.

Anjali also explores how AI is set to transform DevOps, not just by automating repetitive tasks but by becoming an integral “team member” that enhances collaboration across security, IT and development. She stresses that AI-powered DevOps tools will become essential, with predictive insights, adaptive infrastructure provisioning, and automated security becoming the norm. However, the biggest challenge lies in shifting organizational culture—companies must embrace AI while maintaining strong governance to ensure trust and security.

February 6, 2025

Celebrating 10 Years of the North American DevOps Group with Jim Schild

Jim Schild, founder and head of community at the North American DevOps Group (NADOG), discusses the journey of building a DevOps community from the ground up. Originally an English literature major with aspirations of teaching, Jim found himself in the tech industry, eventually immersing himself in the DevOps space. In 2015, he launched NADOG as a grassroots initiative to foster real connections within the DevOps community. The group has since grown into a key player in North America, emphasizing peer-led knowledge sharing and local networking over traditional marketing-driven events. Jim shares how the community adapted through the pandemic, transitioned back to in-person events, and continues expanding across major cities.

Jim also touches on the evolving definition of DevOps, the rise of platform engineering, and how marketing hype has shaped industry perceptions. Jim reflects on the challenges of maintaining a thriving community, the importance of authenticity in engagement, and his commitment to keeping NADOG a space for real practitioners. Looking ahead, he highlights upcoming events and potential expansion into South Florida, inviting IT leaders and practitioners to get involved.

February 6, 2025

Decentralized Application Development with Mathias Buus Madsen

Mathias Buus Madsen, CEO of Holepunch, joins the conversation to challenge the industry’s heavy reliance on centralized platforms. The tech world increasingly gravitates toward centralization and large data centers, and Mathias advocates for a return to the decentralized roots of the internet. He highlights how centralized systems, despite their efficiencies, introduce limitations like privacy concerns, security vulnerabilities, and increased operational costs.

The conversation also explores the practicalities of decentralized development. Peer-to-peer systems, while requiring a mindset shift from traditional server-based models, bring unique advantages like built-in encryption, fewer operational dependencies, and enhanced scalability. By rethinking how applications are built and deployed, developers can simplify processes, reduce vulnerabilities, and create systems that are inherently more secure and efficient. This decentralized approach aligns with a growing demand for privacy and autonomy in today’s interconnected world.

January 28, 2025

Coding AIs in Large-Scale Software Engineering with Scott Dietzen

Scott Dietzen, CEO of Augment Code, discusses how artificial intelligence (AI) is reshaping large-scale software engineering. He shares his background in machine learning and his experience working with massive, complex codebases, highlighting the challenges engineers face when maintaining and updating them. Then he explains how AI can help developers navigate these challenges by understanding the structure of their code, reducing errors, and making large projects more manageable.

Dietzen also talks about the broader impact of AI on software development, sharing examples of how companies are using AI to onboard engineers more efficiently and speed up complex migrations. Lastly, he emphasizes how AI can help reduce technical debt, improve software reliability, and support long-term maintenance.

January 28, 2025

Category: Video Interviews