- Published on
Building With AI After Fifteen Years in Web Technology
What fifteen years across agency work, Tableau, Salesforce, and Veeva taught me about evaluating AI-assisted development in practice.
Why This Blog
I have spent more than fifteen years watching new technology arrive with promises to transform how teams build for the web. Some changes earned a permanent place in the stack. Others created a burst of activity and quietly disappeared. AI-assisted development feels consequential, but the interesting question is not whether the tools are impressive. It is whether they help us make better systems and better decisions.
This blog is where I share that practitioner view: what I am using in real projects, where it creates leverage, where it introduces risk, and how experience changes the way I evaluate it.
Where I'm Coming From
My career has moved through agency delivery, product and platform work at Tableau and Salesforce, enterprise systems at Veeva, and now independent consulting. Across those environments I have worked beside engineers, designers, product managers, marketers, and executives who needed technology decisions to survive contact with real organizations.
That cross-functional position is a credential, not a disclaimer. I can prototype, debug, and work directly in code, but I also know that shipping code is only part of the job. Architecture, governance, usability, organizational constraints, and the cost of future change matter just as much. AI makes implementation faster; it does not remove the need for judgment.
Why Now
AI has compressed the distance between an idea and a working implementation. That is useful, but speed changes the risk profile. Teams can now produce plausible code faster than they can understand, review, or maintain it. The durable advantage will not come from generating the most code. It will come from pairing faster execution with clear constraints, strong review, and an architecture that can absorb change.
That is the transition I want to examine: not AI as spectacle, but AI as part of a professional delivery system.
What's Next
I will write about workflows, tools, architecture, and case studies grounded in actual projects. Some posts will be time-bound check-ins because the tools are moving quickly. Others will focus on principles that should outlast the current generation of products.
The goal is straightforward: show the work, take positions, and share what changes when experienced practitioners can build more directly without giving up the judgment that experience earned.