Blog
February 10, 2026·4 min
What I look for when choosing a stack
I pick tools for the next six months of shipping, not for an imaginary perfect architecture.
Stack decisions get romantic fast. It is easy to optimize for elegance and end up with a setup that is hard to deploy, hard to hire for, or hard to debug under deadline.
My filter is boring on purpose: can I ship a vertical slice this week, can I host it without ceremony, and can I explain the pieces to another engineer in ten minutes? Next.js, TypeScript, and a small backend or serverless surface usually clear that bar for the work I do.
I also care about escape hatches. If the project needs queues, vector search, or a heavier Python service later, I want seams that can grow without a full rewrite.
The best stack is the one that stays out of the way while the product gets clearer. Novelty is optional. Clarity is not.