I've been writing software professionally for over 12 years. In that time, the biggest shift in how I work hasn't been a new framework or language — it's been AI tooling.
Not in the way you might expect. I'm not generating entire applications with a prompt. I'm using AI to compress the parts of development that used to eat the most time without adding the most value.
What AI actually changes in my workflow
Here's where AI tooling makes the biggest difference in my day-to-day client work:
Research and discovery
When I'm building an integration with an API I haven't used before, the old process was: read the docs, find examples, experiment in a sandbox, figure out the auth flow, handle edge cases. That's easily a half-day just to get oriented.
Now I can get up to speed on a new API in under an hour. I use Claude to summarize documentation, generate initial client code, and identify the gotchas before I hit them.
Boilerplate and scaffolding
Every Laravel project has the same foundational code — migrations, models, form requests, resource controllers. I've written this code hundreds of times. AI generates it in seconds, and I review and adjust it rather than writing it from scratch.
Code review and debugging
When something isn't working and I can't immediately see why, having an AI analyze the code and suggest potential issues is like having a senior developer available 24/7 for rubber duck debugging. It doesn't always get it right, but it usually points me in the right direction faster than staring at the code alone.
Documentation and communication
Writing clear technical documentation, project updates, and client-facing explanations of complex systems used to be a separate task I'd batch at the end of a sprint. Now I generate first drafts as I work and refine them. The output is better because it's written closer to when the work was fresh.
What AI doesn't change
AI doesn't replace the parts of my work that clients are actually paying for:
- Architecture decisions. Knowing whether to use a queue vs. a cron job, when to denormalize data, or how to structure an integration for maintainability — these come from experience, not prompts.
- Understanding business context. AI doesn't know that your warehouse team processes returns differently on Mondays, or that your sales team needs the dashboard to load in under 2 seconds because they demo it on calls.
- Quality judgment. AI-generated code can look right and still be wrong in subtle ways. Knowing what "production-ready" means for a specific client requires human judgment.
- Communication and trust. Clients hire me because they trust me to make good decisions and tell them when something won't work. No AI replaces that relationship.
My current AI stack
- Claude for code generation, analysis, and technical writing
- Cursor as my primary editor with AI-assisted autocomplete and chat
- GitHub Copilot for inline suggestions during focused coding sessions
I switch between these depending on the task. Claude is best for complex reasoning and longer-form work. Cursor is best for interactive development. Copilot is best for small, repetitive completions.
The honest trade-off
AI makes me faster, but it also raises the bar for what clients can expect. When boilerplate takes minutes instead of hours, there's more time for the work that actually matters — better error handling, more thorough testing, cleaner code structure, and more thoughtful UX.
The developers who will thrive are the ones who use AI to do better work, not just more work.
What this means for my clients
Practically, it means:
- Projects move faster. Not because I'm cutting corners, but because the mechanical parts of development take less time.
- More time on what matters. I spend a higher percentage of each project on architecture, testing, and polish — the things that determine whether software holds up over time.
- Lower cost for the same quality. When I can do in 30 hours what used to take 40, clients benefit directly.
I'm transparent about using AI tools because I think it's the right thing to do. It doesn't diminish the work — it amplifies it.
Want to see what AI-augmented development can do for your project? Check out my web and app development services or get in touch to discuss your next build.