Now is the time to rethink engineering productivity
Transformations at LinearB for the AI-driven era
Are your teams feeling the intense pressure to "produce more" in an era increasingly dominated by AI?
Join hosts Ben Lloyd Pearson and Dan Lines as they unpack a major shift in how engineering organizations must now approach productivity. Dan reveals the urgent challenges he hears directly from CTOs and VPs, who are grappling with how to define their AI strategy for genuine productivity gain, accurately measure its true impact, and understand the resulting implications for their workforce.
In this discussion, Ben and Dan explore why traditional software engineering intelligence (simply having metrics and information) is no longer sufficient in 2025. Together, they explore "productivity's" nuanced meaning and discover how organizations can shift from passive data observation to an active improvement mindset, and get a look at what defines a developer productivity insights platform.
"Just having software engineering intelligence... that used to be enough, but that's no longer the case... the difference is we now need to take that information and we need to apply immediate actions." —Dan Lines
The Download
The Download is where engineering meets new research, bold opinions, and practical tactics. ✨
1. Dev Interrupted gets a new coat of paint 🎨
Long-time fans of Dev Interrupted will notice some fresh visuals across our platforms! We’re stepping into 2025 in style with a new color theme and logo. You’ll see our primary sponsor LinearB on some of this new branding, as well. This is a good reminder to follow Dev Interrupted on LinkedIn if you’re not already, and make sure you’re subscribed to the pod on your favorite podcast app.
Do: Let us know what you think in the comments!
2. Why your AI coding assistant keeps messing up 🤖
Fractional CTO Pete Hodgson’s latest primer dives into the world of agentic AI coding, arguing that the real question isn’t whether AI can write good code, but rather when it shines and when it flops. By introducing a “Constraint-Context Matrix,” he outlines how to effectively harness AI’s strengths while avoiding its pitfalls. Think of AI as a senior developer with the decision-making skills of a newbie: guidance is essential!
Read: Why Your AI Coding Assistant Keeps Doing It Wrong, and How To Fix It | Pete Hodgson
3. The enduring power of pen and paper ✍️
Developer Juha-Matti Santala reminds us that sometimes the best ideas come away from the screen. His recent article reflects on the use of pen and paper for capturing creative thoughts, arguing that a physical notebook can foster deeper thinking and long-term knowledge retention. Whether it’s on a walk or a cozy couch, stepping away from the IDE can lead to breakthroughs.
Read: As a developer, my most important tools are a pen and a notebook
Don’t miss tomorrow’s workshop about agentic AI adoption 👥💭 (sponsored)
Tomorrow, join host Andrew Zigler for a live 35-minute virtual panel on the future of AI in the SDLC. The webinar will feature past Dev Interrupted guests Birgitta Böckeler from Thoughtworks and Adnan Ijaz from Amazon Q, along with Suzie Prince from Atlassian. We’ll dive into how leading teams are pushing the boundaries of Copilot: experimenting with agentic AI, measuring real impact, and driving meaningful DevEx gains.
Can’t make it live? No worries! Register to get the recording and a comprehensive guide. Don’t miss out, see you tomorrow.
4. Shadow development and the unsanctioned reality of engineering 💻
Tech journalist Kelli Korducki’s recent LeadDev article sheds light on the rise of “shadow development,” where engineers use unsanctioned tools to bypass cumbersome processes. This often stems from a disconnect between leadership and the team, leading to creative workarounds that can pose cybersecurity risks. As a phenomenon, the realm of shadow IT is not new, but it continues to take evolving forms. What we’ve learned from this form is that leaders must either support the tools developers need or risk them adopting their own, often without oversight.
Read: The problem with shadow development
Just listened to this episode. I think we as an industry are missing an opportunity to think about engineering productivity. If we rethink about iterative development, it's very much about idea validation. Not just in the feature shipped, but also in the approach taken. It's why we run spikes in the team. As the cost of code approaches zero it becomes cheaper for us to validate more ideas/options. Ideas in the past that may have been to risky to contemplate, because of the time sink, now become less costly. I think there is a measure of productivity around ideas validated that we're missing. I don't think cycle time encompasses this, as it may be optional scope that ends up getting delivered because we find ways to do it cheaper than if we didn't have AI assistance throughout the team.
Trying to keep it at the team level as UI tools, product definition tools are just as impactful as the Dev tools