Union Pacific Railroad — Paige Jensen

Overview

One view into the health of a railroad's entire IT infrastructure

Unified Systems Management (USM) is an IT infrastructure-management platform for Union Pacific Railroad, one of America's largest freight railroads. Its job: surface the health of UPRR's sprawling systems in real time, alert the right team the moment something goes wrong, and get the right level of information to the right person so incidents get resolved fast.

The goals were concrete: reduce mean time to detection and recovery (driving higher availability), mitigate risk with real-time access to critical information, route alerts to the right support team based on root cause, and give every role the right altitude of data — from a single executive health score down to an engineer tracing a failing service.

USM Macro Health Dashboard — enterprise system health in one view

The USM Macro Health Dashboard — the entire enterprise's system health in a single, drill-downable view.

The Problem

A platform people didn't trust, drowning in noise

Research surfaced a blunt user truth: "Users don't trust the platform because they don't know where the data comes from — and they're overwhelmed by the spammy nature of the alerts." In practice, an outage flooded many engineers' inboxes at once with no clear ownership — they had to sort through a pile of alert emails just to understand what was actually down, with no visibility into whether anyone else was already working to resolve it. The platform wasn't just a visibility gap; it was producing active confusion at exactly the wrong moment.

And the underlying data, where it existed, was often unusable — complex service call-chains, for example, were rendered in a way that was virtually impossible to read.

Example of one alert email — users faced a spammy inbox of these with no clear ownership.

How service call chains were visualized in USM at the time — dense and unreadable — How complex service call-chains were visualized at the time — virtually unreadable.

Dedicated alerting interviews made the fixes clear: the system needed to know when outages were planned — deployments, blackouts, grey-outs — so it wouldn't cry wolf; it needed dynamic thresholds rather than static ones; and notifications needed team-level enrollment, customizable down to the individual, to ensure accuracy, reduce noise, and get the right alert to the right person.

My Role

Lead designer, embedded with the client end to end

I led the design of USM and worked embedded with Union Pacific throughout — running the interviews and workshops, synthesizing the research, prototyping, and steering the executive reviews myself rather than through an account team. That direct access kept the designs grounded in how UPRR actually operated. Across the engagement I designed the platform's core surfaces — the Macro Health dashboard, the Domain Owner view, the Business Features (view and manage), and the alerting experience — and partnered with product and engineering to turn it all into a prioritized, buildable roadmap.

The Process

Interviews → synthesis & prototyping → executive review

User interviews. I led interviews and workshops with internal stakeholders and the engineers who actually used the platform, surfacing where trust broke down and where the data fell short.

Synthesis & prototyping. I synthesized the research into design directions and prototyped fast — including on paper — to pressure-test ideas before investing in pixels. A competitive and comparative analysis of what already existed in the market informed the decisions.

Executive review. I presented two concepts to the executives and ran a structured decision-making session: each stakeholder used voting dots to mark the pieces of each concept they preferred, and explained why. Overnight, I synthesized those insights into a single unified concept and presented it back to the team to confirm alignment.

Paper prototyping the USM dashboards — Synthesis & prototyping — working through the structure on paper first.

Workshop session with UPRR stakeholders — Workshops and interviews with UPRR stakeholders.

Executives voting on concepts with dots — Executive review — voting on the pieces of each concept they preferred.

Prioritization

From insight to a buildable roadmap

I organized the raw feedback into quotes and insights by team and theme — alongside screenshots of the reporting dashboards the train team relied on day to day — then worked with the client to prioritize: plotting user stories on impact versus effort, partnering with product to write and organize the stories, and working with the technologists to land on the best solutions to users' pain points.

Example of a reporting dashboard the train team used at the time

Example of current reporting dashboards.

Wireframes

Reframing the unreadable into something usable

Before high fidelity, I worked the flows as wireframes — reframing those dense, unreadable call-chains into something a person could actually use. Here's the prototype in motion:

The Business Features wireframe prototype.

The Designs

One platform, three altitudes of the same data

Collapsing everything into one view would have caused alert fatigue and made the tool useless for everyone. So the platform served distinct roles with distinct views of the same underlying data — mirroring how UPRR itself was organized, which made adoption natural.

Macro Health

Executive

The overall system health of the enterprise in one view — with drill-in and diagnostics on the very next click, and clear prompts for where to intervene.

Domain Owner

Manager

A one-stop shop for a domain's health and the jump-off points for diagnostics — so managers can ensure operational efficiency across their area of ownership.

Business Features

Engineer & Architect

View and manage the lowest-level transaction patterns and service call-chains — so engineers can trace root cause and architects can see how everything connects.

Here's how those surfaces came together in high fidelity.

Domain Owner — a one-stop view of a domain's health, with jump-off points for diagnostics.

Business Features (View) — explore how services connect and trace a failure across the call-chain.

Business Features (Manage) — define and manage the lowest-level transaction patterns.

"I would love to understand how you come up with all of this because it is fantastic."
— Arun Giri, AVP Lead Engineer, Union Pacific

Outcomes & Impact

Work that earned its place — and an extended engagement

The executive and product teams were thrilled with both the work and the process behind the Macro Health dashboard — and the engagement was extended on the strength of it. The product genuinely worked: clearer ownership and prioritized, trustworthy alerts meant faster root-cause identification and less downtime.

15% improvement in mean time to recovery of outages

Designed the full platform — Macro Health, Domain Owner, Business Features & Alerting

Engagement extended on the strength of the work and process

One platform serving executives, domain managers, engineers & architects

"I love this process and I think you hit it out of the park."
— Matthew Kasselman, VP of IT Development, Union Pacific

Union Pacific Railroad — Unified Systems Management