The CI/CD parallel is a good one, because it reveals the same flawed architecture. But calling it a "limited, transactional API" assumes the backend is even built for it. In my experience, the console itself is often a thick client monstrosity talking to a jumble of services. Adding a clean API for mobile actions would require a foundational rewrite they've avoided for a decade.
So it's not just built for the procurement checklist. It's built *on top of* a system that was never designed for real-time, external transactions. The mobile app's uselessness is just the symptom of that deeper technical debt.
Anecdotes aren't data.
"Afterthought" is generous. Sounds like the mobile team was handed a list of UI components to render, not an actual workflow to enable.
You're right about the VPN part. The moment a tool adds friction to the critical path, it's anti-tatctic. If I can't acknowledge or assign, it's not a monitoring app, it's a notification spambot with extra steps.
The real kicker is they'll call it a feature. It's a liability you pay for.
That "notification spambot with extra steps" description is spot on. It feels like getting a text that just says "there's a problem" with no way to reply.
Does this happen because the mobile team is siloed from the engineers who actually build the console's actions? Like, they're given a design spec to build a "viewer" but never talk to the team managing alerts.