Bhavin Sheth

Really appreciated how you connected the architecture with a working app. The CRUSH and RADOS explanations finally clicked for me because of the practical examples.

Really like the request vs response handler breakdown. Intercepting tool calls before they run feels like one of those small details that makes AI agents much safer in production.

This is the part many teams are learning the hard way: AI speeds up writing code, but not understanding it. The bottleneck just moves from coding to review, testing, and ownership.

Most people focus on design first, but if search engines and AI tools can't understand the content structure, the best UI in the world won't help much. Good insight.

I build my website allinonetools.net exactly that same idea Open → Do → Close.

Interesting checklist. LLM-generated SQL feels easy in demos, but production safety is a completely different problem.

Loved that you included what broke along the way instead of only the final architecture. The JWT filter → exception handling lesson is one of those things you only remember after hitting it once.

Really liked that you showed the mistakes too — that’s the part most tutorials skip. Building flat first and modularizing later is such an underrated way to actually understand Terraform.

Feels like we didn’t remove friction, we just moved it to “verification.” If users have to double-check every output, it’s not solving UX. For me, if it breaks predictability, it’s not worth shipping.

This is actually one of those guides you wish you had before deploying 😅 The small details (PORT, Prisma, routing) are what usually cost hours.