Good technology doesn't have to be bigger than the problem it's solving.

Start simple

If a problem can be solved with straightforward, well-tested technology, that's where I start — before reaching for AI at all. Simple solutions are honest ones: easier to trust, cheaper to run, and far less likely to let you down.

Right-sized, not oversized

Not every task needs a massive AI model, and not every task should avoid one, either. I try to match the size of the tool to the size of the problem — sometimes that's something small and efficient, and sometimes it's genuinely a large, powerful model. What matters to me is that it's a decision, not a default.