To start, could you share the kinds of problems you’re focused on solving right now?
As a software engineer in the year 2026, my focus has shifted significantly from "building the piping" to move data around to more of a supervisory role for increasingly competent AI agents.
I now have more of a product focus, trying to solve issues such as scaling workflows efficiently or utilizing AI to crack business problems.
Looking back, what key decisions or experiences put you on the path from general software development into full-stack, cloud-native work in the computer software industry?
I started my career as a backend Java developer. It was a valuable experience to understand how to design robust, fault-tolerant systems. I made early mistakes and learned from them. Now, I can guide AI agents through the same learning path.
The key to success now is being open to learning new skills all the time. The focus is now less on how to implement complicated workflows and more on how to use the tools at your disposal to do so.
You’ve highlighted the rise of v0 AI-assisted tools; from your hands-on use, what criteria and guardrails have helped you adopt them effectively without compromising code quality or team velocity?
Agents need guidance. Just like you'd guide a junior developer with documentation, you can provide valuable context to an AI agent using markdown files. It also helps when your codebase has well-established and repeatable patterns to perform common tasks, such as tables, forms, workflows, etc.
Once a v0-style prototype shows promise, what is your playbook for hardening it for production—especially around automated testing, CI/CD, and code quality practices that have worked for you with tools like Jenkins and Git?
Any v0-style output still needs a manual review by a developer. Things to look out for include:
- performance criteria,
- adherence to best practices already established, and
- the ability to scale a solution to support arbitrarily large inputs.
On the front end, how do you leverage Next.js, React Server Components, TypeScript, and Tailwind CSS to deliver performance and accessibility from day one without slowing down iteration?
TypeScript has always had a rich library ecosystem to choose from, which allows developers to get productive fairly quickly. This has now been turbocharged by the introduction of more AI-friendly toolsets, such as skills, CLIs, and MCP servers.
To close, what cross-functional practice—spanning product collaboration and stakeholder management—has most improved delivery speed without sacrificing reliability, and how did you implement it with your teams?
Good communication is key at all times. Co-location of product and developer teams is essential. Nothing should be lost in translation. With AI reducing the gap between prototype or design and production, it becomes extremely important that both your development team and product team are in sync at all times.