Streamlining Infrastructure Mapping for Metro-X

When an outage hits, every minute costs money. DevOps teams at Amdocs were spending 40% of incident response time just figuring out what's connected to what. I designed the system that gave them that answer in seconds.

👤 Role

Lead UX Designer

💼 Industry

Enterprise B2B

📅 Duration

6 month

Metro-X
Metro-X
Metro-X

Overview

Overview

Context

Metro-X is an enterprise infrastructure management platform designed for DevOps and infrastructure teams.

Teams rely on Metro-X : to manage complex ecosystems composed of environments, servers, applications, databases, load balancers, and more. Fast, accurate understanding of how these entities are connected is critical for troubleshooting, change management, and maintaining uptime.

My Role

I was working in a design team.

My Role

I was working in a design team.

I collaborated with

1 PM, 2 backend engineers, 1 frontend engineer, SRE advisors from client teams


What I owned:

  • Research strategy and stakeholder alignment

  • Information architecture for the topology system

  • Interaction model and progressive disclosure framework

  • Design system extensions (12 new infrastructure icons)

  • Phased release strategy (negotiated with PM and engineering)

Problem

Enterprise DevOps teams at Amdocs were burning 40% of incident response time manually mapping infrastructure dependencies through spreadsheets and static diagrams. Critical outages stretched 5+ hours due to inability to quickly identify impacted systems across complex telecom environments.

Solution

I designed an phased, interactive, real-time topology visualization that automatically maps relationships between servers, applications, and databases and other entities of infrastructure ecosystem. Used progressive disclosure approach to prevent cognitive overload while enabling deep-dive analysis when needed.

Impact

35–50%
  1. Faster incident resolution

100+ hours/month
100+ hours/month
100+ hours/month
  1. productivity

20–30%
20–30%
20–30%
  1. better SLA compliance

Scroll down to dive deeper

Research & Discovery

User Research Strategy

I combined quantitative performance analysis with qualitative workflow observation to understand where teams were actually losing time.

User Research Approach
  • 12 in-depth interviews with DevOps engineers, SREs, and infrastructure managers

  • 5 contextual inquiries during live incident response

  • Workflow analysis across enterprise client environments

Research & Discovery

User Research Strategy

I combined quantitative performance analysis with qualitative workflow observation to understand where teams were actually losing time.

User Research Approach
  • 12 in-depth interviews with DevOps engineers, SREs, and infrastructure managers

  • 5 contextual inquiries during live incident response

  • Workflow analysis across enterprise client environments

Navigation chaos

Inefficient Workflows

Irrelevant Content

Collaboration Gaps

Existing process
Existing process
Existing process

Image 2 : Screenshots of current tools and systems in place and also compared topology structures (star, mesh, tree, bus, hybrid) to assess readability, scalability, and cognitive load for large graphs.

Image 3 : Explored ecosystem visualization in Datadog, Dynatrace, AppDynamics, Grafana to derive interaction and legend conventions, and to set performance expectations.

Image 4 : Current Workflow of teams

Key Research Insights

#2
Manual Overhead

Teams spent 5-8 hours weekly maintaining fragmented documentation. Information outdated within days of creation

#1
Visibility Gap

"We can see all our entities but have no single source of truth for relationships."

#3
Incident Cascade Effect

Without relationship visibility, small issues became major outages

#4
Scaling Impossibility

Manual mapping broke down exponentially as infrastructure grew

#4
Scaling Impossibility

Manual mapping broke down exponentially as infrastructure grew

My Design Approach

Mapping Entry points and user flow
  • Mapped primary entry from the dashboard to affected component view in 2–3 clicks; optimized “time to first signal” with lazy loading for details and quick filters for scope.

  • Orchestrated drill‑downs: component → immediate dependencies → blast radius preview → detailed node panel with actions.

My Design Approach

Mapping Entry points and user flow
  • Mapped primary entry from the dashboard to affected component view in 2–3 clicks; optimized “time to first signal” with lazy loading for details and quick filters for scope.

  • Orchestrated drill‑downs: component → immediate dependencies → blast radius preview → detailed node panel with actions.

Entry points
Entry points
Entry points

Image 5 : Entry points from dashboard home.

New User Flow
New User Flow
New User Flow

Image 6 : After identifying key entry points, I mapped the new user flow to illustrate how users would seamlessly access and utilize the topology feature for troubleshooting and dependency mapping.

Mapping Entities and Relationships

To lay the foundation for the topology feature, I systematically mapped out all key entities and their relationships within the Metro-X ecosystem relevant to users’ troubleshooting and operational workflows.

Mapping Entities and Relationships

To lay the foundation for the topology feature, I systematically mapped out all key entities and their relationships within the Metro-X ecosystem relevant to users’ troubleshooting and operational workflows.

Image 7 : E-R Mapping and Feasible Key Information

Design Reviews and Feasibility Loops
  • Weekly triads with product and engineering to have design reviews, feasibility checks and scope check.

  • Maintained a decision log capturing scope, non‑goals, and trade‑offs to prevent scope creep.

Design Reviews and Feasibility Loops
  • Weekly triads with product and engineering to have design reviews, feasibility checks and scope check.

  • Maintained a decision log capturing scope, non‑goals, and trade‑offs to prevent scope creep.

Design System Integration
  • Leveraged existing iconography where possible

  • Created 12 new icons for infrastructure entities

  • Established color coding for relationship types

Design System Integration
  • Leveraged existing iconography where possible

  • Created 12 new icons for infrastructure entities

  • Established color coding for relationship types

Mapping Entities and Relationships
Mapping Entities and Relationships

Image 8 : Metro-X Design System

Design Decisions

Decision #1: Context-Driven Progressive Disclosure

The Design: Default view shows only first-degree dependencies. Zoom or click to reveal downstream relationships.

Strategic Rationale:

  • Our research showed 80% of troubleshooting involved 1-2 degrees of separation

  • Datadog and Dynatrace show everything at once—our users found this overwhelming during incidents (high-stress moments = low cognitive bandwidth)

  • Engineering constraint: rendering 100+ nodes at once caused 3-second lag. Progressive loading kept interaction under 200ms.

Design Decisions

Decision #1: Context-Driven Progressive Disclosure

The Design: Default view shows only first-degree dependencies. Zoom or click to reveal downstream relationships.

Strategic Rationale:

  • Our research showed 80% of troubleshooting involved 1-2 degrees of separation

  • Datadog and Dynatrace show everything at once—our users found this overwhelming during incidents (high-stress moments = low cognitive bandwidth)

  • Engineering constraint: rendering 100+ nodes at once caused 3-second lag. Progressive loading kept interaction under 200ms.

Decision #2: Interactive Exploration Architecture

The Design:

  • Zoom/Pan: Navigate large topologies intuitively

  • Hover Details: Quick information without context switching.

  • Actions: Direct actions can be taken on component/entities via 3 dots.

  • Side Panel for more details: Deep dive details for component/entity clicked.

  • Export Options: Support documentation workflows

Strategic Rationale:

  • Eliminates tool switching during incidents

  • Supports both analysis and action workflows

  • Enables sharing and offline analysis

Decision #2: Interactive Exploration Architecture

The Design:

  • Zoom/Pan: Navigate large topologies intuitively

  • Hover Details: Quick information without context switching.

  • Actions: Direct actions can be taken on component/entities via 3 dots.

  • Side Panel for more details: Deep dive details for component/entity clicked.

  • Export Options: Support documentation workflows

Strategic Rationale:

  • Eliminates tool switching during incidents

  • Supports both analysis and action workflows

  • Enables sharing and offline analysis

Decision #3: Integration-First Approach

The Design: Built within existing Metro-X platform with familiar navigation patterns.

Strategic Rationale:

  • Leveraged existing user mental models

  • Reduced training overhead for adoption

  • Positioned topology as core platform capability

Decision #3: Integration-First Approach

The Design: Built within existing Metro-X platform with familiar navigation patterns.

Strategic Rationale:

  • Leveraged existing user mental models

  • Reduced training overhead for adoption

  • Positioned topology as core platform capability

Validation & Testing

Strategic Testing Approach

Conducted task-based testing with real incident scenarios using enterprise infrastructure data to validate design decisions before rollout.

Testing Framework
  • 14 participants: DevOps engineers, SREs, infrastructure managers

  • Real scenarios: Database failures, dependency mapping, documentation workflows

  • Enterprise data: Complex topologies with 100+ interconnected components

Pre-Launch Results

Metric

Target

Achieved

Impact

Task Success Rate

80%

91%

Intuitive without training

Dependency Mapping

<10 min

5 min

83% faster than manual

User Satisfaction

4.0/5

4.6/5

High adoption confidence

Validation & Testing

Strategic Testing Approach

Conducted task-based testing with real incident scenarios using enterprise infrastructure data to validate design decisions before rollout.

Testing Framework
  • 14 participants: DevOps engineers, SREs, infrastructure managers

  • Real scenarios: Database failures, dependency mapping, documentation workflows

  • Enterprise data: Complex topologies with 100+ interconnected components

Pre-Launch Results

Metric

Target

Achieved

Impact

Task Success Rate

80%

91%

Intuitive without training

Dependency Mapping

<10 min

5 min

83% faster than manual

User Satisfaction

4.0/5

4.6/5

High adoption confidence

Key Learnings

What I learnt?

  • Start with highest-impact use case, not most comprehensive solution

  • Technical constraints often reveal more innovative design opportunities

  • Enterprise users value reliability and performance over visual novelty

  • Cross-functional alignment requires continuous communication, not just kickoff meetings

Challenges That Made Me Stronger

  • Complexity Management: Learned to balance comprehensive functionality with usable simplicity

  • Performance Optimization: Gained expertise in designing for large-scale data visualization


Key Learnings

What I learnt?

  • Start with highest-impact use case, not most comprehensive solution

  • Technical constraints often reveal more innovative design opportunities

  • Enterprise users value reliability and performance over visual novelty

  • Cross-functional alignment requires continuous communication, not just kickoff meetings

Challenges That Made Me Stronger

  • Complexity Management: Learned to balance comprehensive functionality with usable simplicity

  • Performance Optimization: Gained expertise in designing for large-scale data visualization


#2
Manual Overhead

Teams spent 5-8 hours weekly maintaining fragmented documentation. Information outdated within days of creation

#3
Incident Cascade Effect

Without relationship visibility, small issues became major outages

#4
Scaling Impossibility

→ Manual mapping broke down exponentially as infrastructure grew

"We can see all our entities but have no single source of truth for relationships."

#1
Visibility Gap

See also

© Made by Meghna Aggarwal with creativity, lots of ctrl+z, and coffee. ⋆.˚☕︎

⏱ Toronto, ON •

16:05

© Made by Meghna Aggarwal with creativity, lots of ctrl+z, and coffee. ⋆.˚☕︎

⏱ Toronto, ON •

16:05

© Made by Meghna Aggarwal with creativity, lots of ctrl+z, and coffee. ⋆.˚☕︎

⏱ Toronto, ON •

16:05

Create a free website with Framer, the website builder loved by startups, designers and agencies.