April 16, 2026
Subject: Quick follow-up — Claude Code clarification
Thanks for the time today. The 90-95% NPS score you mentioned stuck with me — that's the kind of metric that shows you're solving a real problem, not just building features.
I didn't land this well when you asked about developers reviewing Claude-generated code. I went straight to the failure mode I've seen elsewhere. What I should have led with: I'm building with Claude Code every day as my primary engineering tool right now. Not evaluating it — shipping with it. Production features, architecture decisions, code review discipline built into the workflow.
I mentioned rail billing and dementia detection as areas I've been exploring. What I should have said more clearly: Dimer is the role I'd actively choose. Post-hospital care, Claude Code in the stack, Series A, hybrid team with senior Polish engineers I'd partner with rather than replace.
Scale context I didn't land: the platform I built at Credentia is still running today, processed over a million certifications across 35 states, bootstrapped, profitable from year one, no outside capital. Happy to connect you with three people who saw me operate across a full product cycle — a former co-founder, my operations lead, and a board member. Let me know if that would be useful and I'll make the introductions. Either way, appreciated the conversation.
Kevin