A Federal Reserve audit exposed unreliable inventory controls across a global network estate. I turned the fix into a platform — built around trusted asset state, governed lifecycle, safer automation, and executive visibility.
The estate supported $100B in daily transaction notional. Stale DNS records, conflicting IP allocations, and fragmented device state were not just admin mess — they affected change safety, incident response, audit posture, and automation. The audit gave a deadline. I used it as a reason to build something that would last.
Each screen frames the product around a single outcome the buyer cares about. No feature lists, no dense walkthroughs.
Screenshots from the actual interface. Strong hierarchy, restrained colour, explicit workflow cues — so technical detail stays readable under real operational conditions.
I wanted the dashboard to compress a massive estate into a few decisive metrics: total assets, lifecycle verification, monitoring coverage, and critical exceptions. If an executive can't understand the posture in ten seconds, the dashboard has failed.
Severity, count, business risk, and next action all live together. The side panel gives context without forcing a full page change. I spent time making sure the triage flow felt like a single, coherent decision rather than a series of disconnected screens.
Complex asset history turned into a readable timeline of state transitions — supporting events, ownership, and control status. If someone asks 'what happened to this device and when?', the answer should be immediate.
Passed validations, failed validations, block reasons, and required actions are explicit. Teams can see why something is blocked and what to do about it. No mystery, no tribal knowledge.
Before-and-after movement, structured for executive review, programme ownership, and audit discussion. I wanted the reports to tell a story — not just dump data.
Unreliable network inventory controls turned into a production-grade platform — backed by 99.9% uptime, enforceable lifecycle controls, and disaster-recovery readiness. The hardest part wasn't the engineering. It was making complex state legible without oversimplifying what operators actually need.