agentic SDLC Archives - Hoola Hoop - Executive Coaching for CEOs, CTOs & Boards

From SDLC to Agentic SDLC.

I’ve lived through a lot of process evolutions. The move to agentic development is different in kind, not just degree. It’s changing what it means to lead an engineering organization altogether.

CTOs aren’t asking “should we use AI?” anymore. That debate is over. They’re asking: how do we rebuild our development process around it and how do I need to lead differently?

This article is my attempt to answer that.

What Traditional SDLC Was Built For

The Software Development Lifecycle: requirements, design, development, testing, deployment, feedback was architected around a fundamental constraint: humans are the only ones who can do the work.

That constraint shaped everything. Sequential handoffs existed because one person or team needed to finish before the next could start. Sprints existed to timebox human capacity and velocity. QA came after development because writing tests and writing code at the same time was too expensive. Code reviews were async because engineers couldn’t be in two places at once.

We built an entire system of process around human throughput limitations. And for decades, it worked.

What’s Changed

Agentic AI tools like Claude Code, Cursor, Copilot, and Devin are changing more than individual developer speed. They’re collapsing and reinventing the handoffs between stages entirely. Here’s what I’m seeing as a CTO and in the organizations we advise:

Requirements

What used to take weeks of grooming sessions is being drafted, refined, and structured by AI in hours. Engineers are parsing business goals and generating user stories directly but not waiting for a PM to hand off a spec. PMs who embrace this shift move from writing specs to shaping strategy, but those who don’t will find their role increasingly redundant.

Design & Architecture

What used to require a senior architect and days of whiteboarding can now produce multiple candidate architectures with trade-off analysis in a single session. The human job shifts from creating the design to evaluating and deciding between options.

Development

Developers are increasingly writing intent and describing what they want to build, while agents scaffold, implement, and iterate. The developer reviews, steers, and catches failure modes. They’re not gone from the process, but operating at a higher altitude. And frankly, I think that makes them more valuable, not less.

Testing

Testing is no longer a phase that happens after the code is written. Agents generate and run tests during development, flag regressions in real time, and maintain test coverage as the codebase evolves. The idea of a separate QA phase is becoming an artifact of the old model.

Deployment & Operations

Agents monitor, alert, and in some cases remediate without waking anyone up at 2am for a failure pattern. The human role shifts from routine incident response to designing the guardrails that determine when and how agents escalate.

What This Means for CTOs

The process implications are real, but the leadership implications are even bigger.

🏗️

Your org chart hasn’t caught up

The traditional split between engineers, PMs, and QA, or between engineers and architects, is eroding fast. If your team structure still reflects a handoff model, you’re adding coordination overhead that no longer serves a purpose. My prediction: Product, Engineering, and Design will ultimately converge into a single unified team.

🎯

The bottleneck has moved

In traditional SDLC, the bottleneck was development throughput and not enough engineers, not enough time. In agentic SDLC, the bottleneck is increasingly decision quality. Are we building the right thing? Is this architecture sound? The humans in the loop need to be excellent at judgment, not just execution.

⚠️

New failure modes require new oversight

Agents fail differently than humans and typically confidently, sometimes quietly, but at scale. They can produce code that looks right, passes tests, yet introduces errors or debt that surface weeks later.

🔲

“Done” is harder to define

In a world where an agent can keep iterating indefinitely, knowing when to stop and ship is a real skill. Scope discipline becomes critically important, not less.

📋

The governance question is unavoidable

What decisions can an agent make autonomously? What requires human approval? What gets logged and audited? Who is the engineer and responsible for the code in production? These aren’t theoretical questions but operational ones every engineering leader needs to answer before they can responsibly scale agentic workflows.

“Ever received a PR from your CEO who’s decided they’re now an engineer via Claude Code? That governance conversation is happening in more organizations than you’d think.”

What Doesn’t Change

I want to be direct about this, because I think there’s a real risk of CTOs either over-correcting or under-responding to this shift.

The human things, specifically the leadership things, DO NOT go away. They become more important.

Your job as a leader is to create the conditions where higher-level thinking can actually happen, which means protecting your best people from being buried in review queues of AI-generated code.

Knowing which problem to solve, and why now, is still yours. AI can accelerate execution with extraordinary speed. It cannot set direction. Strategic judgment, customer empathy, navigating organizational ambiguity, building a team that trusts each other. None of that is on the automation roadmap.

What is changing is the ratio. More of the execution layer is being handled by agents, which means the humans in the loop need to be operating at a higher level not just reviewing code, but shaping outcomes. The exceptional engineers on your team will thrive in this environment. Those who relied primarily on execution throughput will find the transition harder.

The bottleneck in agentic SDLC isn’t development throughput – it’s decision quality. The humans in the loop need to be excellent at judgment, not just execution.

Questions to Sit With

If you’re a CTO thinking through what this means for your organization, these are the questions I’d encourage you to work through:

Does your development process still have handoffs that exist because of human throughput constraints rather than because the handoff itself adds value?
Where is your human review layer, and is it optimized for catching agent failure modes rather than traditional human ones?
Are your engineers spending more time on judgment, direction, and evaluation… or are they still primarily in execution mode?
How are you communicating the implications of agentic development to your CEO and board, and are you bringing them along before the organizational changes become visible?
What does “done” mean in your current process, and does that definition still hold in an environment where iteration is nearly free?

A Final Thought

The CTOs who navigate this transition well aren’t the ones who adopt every new tool fastest. They’re the ones who understand what the tools change about process, about team structure, about where human judgment is irreplaceable and redesign their organizations around that understanding.

That’s not a tooling or process problem. It’s a leadership problem.

Ready to talk about CTO coaching with Leigh?

Book a 30-minute introductory call to explore whether coaching is right for you.

Book a meeting with Leigh →

Leigh Newsome

Partner, Hoola Hoop · CTO Coach

Leigh Newsome is a Partner at Hoola Hoop and a CTO coach with 25 years of experience scaling product and engineering teams. He has worked with a wide range of startups and global enterprises, including Avid, Digidesign, WPP, and Kantar/Millward Brown, and successfully led TargetSpot (backed by Union Square Ventures, Bain Capital Ventures, and CBS) through its acquisition to Radionomy Group (Vivendi). When he’s not coaching CTOs, you’ll find him teaching digital audio to graduate students at NYU, building audio and signal processing applications, or flying fixed-wing aircraft — but never all three at once.

The Agentic SDLC: A CTO’s Guide