Senior Director of Product & Engineering says AI’s biggest impact on product organizations comes from the dismantling of silos
Andrew Knight explains how AI helps product teams break down silos, expand ownership, and prototype faster so people can focus on solving the right problems.
Andrew Knight
Senior Director of Product and Engineering at Cycle Labs
Force Multiplier: AI accelerates product development by helping teams test ideas, write code, and deliver results faster.
Fewer Silos: Combining product and engineering responsibilities creates stronger ownership, less friction, and quicker decisions across teams.
Broader Access: AI coding assistants help nontechnical employees build useful plugins, expanding contribution beyond traditional engineering roles.
Faster Feedback: Rapid AI prototyping lets teams validate customer needs early, replace assumptions, and improve promising features before investment.
Human Judgment: AI should handle repetitive work, while people retain responsibility for discovery, priorities, roadmaps, and user experience.
Andrew Knight is Senior Director of Product and Engineering at Cycle Labs. He's also known as the "Automation Panda" and has a blog under the same name.
We caught up with Andrew to see how he's using AI to break down silos in his organization. Here's what he told us.
AI can be a force multiplier for product teams
My name is Andrew Knight, but I'm better known as the "Automation Panda." My career has focused on building solutions for testing problems. I spent a decade as an SDET. Then, I did a tour of duty in DevRel. Now, I lead a Product and Engineering team at Cycle Labs, where we develop a test automation platform for supply chain systems.
Continue Reading for Free
Create a free account to finish this article, plus get ongoing access to timely insights and practical resources.
Have an account? Log In
Cycle is a test automation platform for warehouse management systems (WMS) and adjacent supply chain systems. We target enterprise customers who need to ensure correct WMS configurations, especially during upgrades. We also target partners to help implement Cycle test projects. Our user base ranges from warehouse managers who don't consider themselves testers to elite SDETs building scalable test suites across multiple sites.
We deliver parts of the platform through multiple offerings. The primary distributable is a Windows desktop app/kit that enables users to write and run automated tests. We also provide libraries of test code to help users jumpstart their test development. And we're building web apps, services, and infrastructure to support test development and test execution.
I started using AI coding tools in earnest about a year ago. Now, my whole team embraces AI to expedite their workflows. We are also building AI capabilities into the Cycle testing platform for our customers and partners. It's exciting!
In the right hands, AI is a force multiplier for product development and engineering.
How AI allows product teams to break down silos
AI enables folks to do more and to break down silos.
Previously, our organizational structure was more traditional; Product Management and Engineering teams operated separately. We have now combined them into one team, breaking down silos and merging responsibilities. We expect our engineers not to be merely "software" engineers but "product" engineers who take ownership of the full scope of problems.
This change is recent, and we are still navigating it, but I'm excited to see where we go. So far, I've seen less friction, more ownership, and faster results.
How AI enables non-technical teams
The dismantling of silos doesn't just apply to product and engineering teams. It goes beyond those.
Here's an example. Cycle provides a special programming language for writing tests. We recently created a plugin architecture that lets testers create new steps for Cycle's language. Our plugin SDK requires Java programming skills, but the structure and patterns are straightforward.
Using AI coding assistants, folks from our operations and services teams are actually able to write their own plugins. Even our former-CEO-turned-Executive-Chairman, who hasn't touched code in at least ten years, successfully built one!
Where AI can be used effectively in product organizations
Everyone on my team has a Claude subscription. Most folks are using it for coding, but some are starting to use Claude Cowork for documents, APIs, and MCP servers.
We also let AI handle experimentation and technical tradeoffs because we're new on this journey, and our product management processes are still less mature than our engineering processes.
I’ve used AI to brainstorm user stories and acceptance criteria that engineers can translate directly into features when the model has the right context and development framework. But I believe humans still need to own discovery, prioritization, roadmap planning, and UX decisions.
And I've used AI to brainstorm user stories and acceptance criteria that engineers can translate directly into features when the model has the right context and development framework.
But I believe humans still need to own discovery, prioritization, roadmap planning, and UX decisions.
How rapid prototyping with AI changes product processes
Rapid experimentation backed by AI coding and prototyping has enabled us to gather feedback on features faster than ever before.
Vibe coding cuts through long cycles of thinking and guessing. The team can code more quickly and in less familiar stacks than before. That means PMs can build and ship prototypes themselves. It also means engineers can turn specifications directly into working code with automated tests. Testers can then extract critical information from both the requirements and the actual code.
So, product managers no longer have to wait for engineers to build things. And we no longer rely on opinions. We build things rapidly, put them in front of customers, and get feedback immediately. From there, we iterate on proven success and pivot away from what doesn't work. And finally, we hand successful prototypes off to engineers.
How AI results in more critical thinking
I've mentioned the time-saving benefits of AI, but what I haven't mentioned is why it matters. The benefit isn't the reclaimed time itself; the benefit is what we can do with it.
As AI takes on much of the grind work of writing specs, writing code, and even writing tests, my team can think more critically about what problems to solve. And that leads to better products.
Why product leaders must push for AI, but use it selectively
AI is more than a faster way to search for information and answer questions. It isn't Google. It’s a tool for improving workflows.
So, here's my advice: Don’t sleep on it. Push the naysayers and sticks-in-the-mud out of the way. Politick and position to make it happen for your organization.
But importantly, don't just throw AI at everything blindly. Use your brain. Build stuff that solves real problems.
Follow along
You can follow Andrew Knight's work on LinkedIn. And check out his blog, Automation Panda.
More expert interviews to come on The CPO Club!
