LabsetReflections on building software that lasts.
I tinker with different stacks and reflect on what holds up, drawing on 15+ years in the field. Follow along, take what helps, and tell me what you'd do differently.
Platform
The ground it all runs on.
Services
The typed API core on top.
Data & AI
Memory and intelligence.
Product
The surfaces people use.
A few principles I keep coming back to
The specific tools are the easy part, and the first to change. What holds steady underneath matters more. These principles come from my own context, so if you'd approach something differently, I'd genuinely like to hear it.
Stack-agnostic
Good foundations translate to any language or framework.
Vendor-agnostic
Choose the providers that fit your context; the foundation holds.
A working prototype
Each foundation becomes a lightweight, working prototype, real code that shows the concept in action.
Transparent decisions
I share the trade-offs, the calls I make, and the ones I would revisit.