It was always about the workflow

I posted a tweet recently and I want to unpack it a bit more because I think it touches on something a lot of us in this space are feeling right now.

The gist: AI is lowering technical barriers and freeing up space for me to focus on what I'm realizing has always been at the center of my work. Not the what we build, but the strategy and efficiency of how we build it. Workflows.

And I don't mean that in a productivity-hack, hustle-culture kind of way. I mean it in the deepest sense of how I think about the craft of making things on the web.

It was always the process

Here's the thing. I've been using design software for about 20 years now. And for as long as I can remember, the moment I loved most wasn't the final export or the launch day screenshot. It was the process that got me there. The figuring-it-out part. The "there has to be a faster way to do this" part.

I've always had a bit of an obsession with finding not just a quicker path, but a more proper path. The kind of approach that holds up when the project gets more complex, or when someone else needs to pick up where you left off. If you've ever dealt with inheriting a messy codebase or a design file with 400 unnamed layers, you know exactly what I'm talking about.

As I moved deeper into code over the last decade, that obsession just expanded. It went from "how do I design this well" to "how do I make this into a real, interactive experience that works for everyone." I got excited about accessibility, interactions, motion, performance. All the things that enhance the experience for anybody using it. And all that to say, those aren't just features or checkboxes. They're workflow decisions. They're about how you set things up from the start so the end result is better for everyone involved.

When the work goes beyond the website

This became especially clear to me on a couple of projects over the years that pushed well beyond what most people think of when they hear "website."

When I worked on Webflow Conf, we pitched building our own registration platform using Webflow, Airtable, Memberstack, and Whalesync instead of relying on typical event registration tools that kept underperforming. The front-end was built in Webflow, but the real substance of the project was the backend logic, the database architecture, the automations, and the internal workflows that powered reporting and attendee engagement across nearly 30,000 registrants. We launched public registration just three weeks after kickoff because the workflow we designed allowed us to move that fast.

A similar thing happened with Partytrick, where we needed to curate and cross-reference over 2,500 collection items in about a month. Doing that manually in the Webflow CMS would have been brutal. So we built a content management workflow using Airtable and Whalesync to live sync everything, and used Airtable scripting for things like bulk image compression. The product itself was great, but the workflow behind it is what made it possible in that timeline.

In both cases, I found myself building something bigger than a website. I was building business functionality, data systems, and automation that drove real value for the team and the client.

The framework obsession

Maybe the clearest example of how deep this goes for me is Mast, the CSS and component framework I've maintained for the last three years.

If you read through the fundamentals in the Mast docs, you'll notice it's not just a collection of classes and components. There's an intentional philosophy underneath it: provide a solid foundation that covers about 80% of any project with default classes and components, but acknowledge that the remaining 20% will always be best handled by your own intuition and experience as a developer. The framework gives you building blocks, not a finished house.

One thing I want to call out is that the underlying thing here is a workflow decision. Mast is opinionated about how you should approach building, not just what tools you should use. Less classes, less elements, less components. Faster site, faster development, faster customization. That philosophy is baked into every decision in the framework, from the 12-column grid to the utility strategy to the naming conventions. It's a way of working, not just a set of styles.

What AI is actually changing for me

Now, here's where this all connects to the current moment.

The more I use Claude Code and other AI tools every single day, the more the technical barriers lower and the more my existing skills are enhanced. And what's left when you strip away the hours of tedious implementation work? The thinking. The decisions about how things should be structured, how data should flow, how teams should collaborate, how the end experience should work for both the people building it and the people using it.

Most recently, I've been exploring building sites and tools with Next.js and Sanity, and it's been especially exciting. That combination strikes the right balance of fundamental functionality, things like real-time collaboration, publishing workflows, databases, and logic that you get out of the box with Sanity, mixed with an open platform where I can define exactly how I want to work and organized content, plus have collaborators manage content and generate pages. I can bake Claude itself into every corner of the platform so it's always ready to help. That's a workflow decision, and it's the kind of thing I find myself thinking about constantly now.

The role we're evolving into

You can sense the mix of excitement and intention in the industry right now. A lot of people are discovering similar possibilities with AI tools. Some see it as a threat, a replacement for the work they've typically done. And I get that feeling.

But here's my take on it: I see it as a way to no longer spend countless hours on tedious work and instead spend the majority of my time focused on how I can guide these tools to create an end experience, both on the front-end and the back-end, that is enjoyable, capable, and performant. The role isn't disappearing. It's shifting toward something I've always cared about most.

I really believe this is the space our work evolves into. Designing and defining the ideal workflows, deciding where to dial up agent involvement and where human judgment is essential, and ultimately creating work that's better because we spent our energy on the decisions that matter most instead of the repetitive stuff that used to eat our days.

The best part? I don't think this is some distant future scenario. It's happening right now for anyone willing to lean into it. If you're in a similar spot, figuring out how AI fits into how you work and what it frees you up to focus on, I'd love to hear about it. This is one of those conversations where I think everyone's perspective adds something valuable.