SmartView data center infrastructure management

Overview

Equinix had shipped an MVP of SmartView — a DCIM platform that lets enterprise customers monitor data center hardware remotely. The MVP wasn't responsive and the patterns across its six monitoring surfaces were inconsistent. My manager tapped me to take it forward. I came in post-MVP, mapped what had been built and decided, then pushed the experience through dashboard, environmental, power, mechanical, electrical, alarms, and alerts — responsive, modular, and handed off at detailed visual design.

The problem

Two structural problems, compounding each other. First: the MVP wasn't responsive, and the primary use case for a DCIM platform is a technician getting an alert at 2am and needing to triage remotely before deciding whether to go in. Mobile wasn't optional.

Second: each of the six data surfaces had been designed in isolation. Environmental, power, mechanical, electrical, alarms, and alerts each had different patterns, different grids, different component conventions. The system felt like six separate tools bolted together. Both problems needed the same fix: a shared component system with a responsive grid at its foundation.

Market context

Data center infrastructure management is a high-stakes category. Environmental monitoring, power draw, mechanical systems, electrical, alarms, and alerts — all real-time, across multiple physical locations, for enterprise customers who measure downtime in revenue loss per minute. Equinix estimated the market opportunity at $20B+.

Competitors — Digital Realty, CenturyLink, NTT, Terremark — were all in progress but none had shipped a full customer-facing DCIM product. That window was the business case for moving quickly. Getting SmartView to production quality before the field consolidated was the goal.

The user

Primary user: data center technician, mid-career, managing multiple IBX footprints across time zones. Core need: operational status without being physically on-site. Core pain: missing alerts, needing on-site presence to diagnose problems, juggling multiple accounts and locations without a unified view.

The design had to work on mobile first, then scale up to desktop operations. The use case where it mattered most — an alert at 2am, a remote triage decision — was a mobile use case. Desktop was for sustained operational monitoring during business hours.

Information architecture

Coming in post-MVP without full documentation meant reconstructing the discovery layer before moving forward. I mapped the existing MVP requirements visually against what had actually been built — a process that surfaced decisions made implicitly during the first phase that needed to be made explicit before the second.

The sitemap built from that pass became the team's shared reference. It hadn't existed before. That document did more for alignment in the first two weeks than anything else — everyone was finally working from the same picture of what the system was and where it was going.

Outcome

Live with Equinix enterprise customers and resellers. The modular section system built here became the pattern for subsequent Equinix product work — a shared component foundation meant that design and development overhead for each new surface dropped significantly compared to the MVP phase, where each section had been built from scratch.