Redesigning Amazon’s Employee Growth Ecosystem
Standardizing framework that scales from a small pilot to a global workforce
Creating a unified onboarding plan framework. I incubated it within the Lift Off program (12K+ users) to test it first, with the long-term vision of scaling it as the standard growth blueprint for all corporate blue badge employees across Amazon.
My role
Lead UX Designer
Timeline
Oct 2025 - Dec 2025
Team
1 designer 1 PM 10 developers
Tools
Figma UserTesting Kiro (Claude model)
Siloed tools, generic content
At Amazon, the onboarding experience depended entirely on which organization you joined. Different business units used completely separate software platforms. To make matters worse, the actual content was "one-size-fits-none."
01
Platform fragmentation
Different orgs used entirely isolated software: Embark for corporate, Lift Off for operations, and ASBX for others. The user experience varied wildly.
02
Lack of customization for individual
A new UX designer would open their plan and find a massive checklist of generalized corporate tasks completely irrelevant to their actual day-to-day discipline or team culture.
Framework over platforms
Instead of trying to redesign three massive software platforms all at once, our team made a strategic decision: design a single, standardized plan framework that could live inside any platform.
Piloting with small number
Partnered with the PM to map a scalable roadmap. We executed a phased rollout using Lift Off (12K+ users) as an incubation sandbox to test and validate patterns with real users before scaling globally. Here are the different phases:
Phase 1
Test within LiftOff
This is the population we are launching! It served as the perfect incubation sandbox to validate our goal-oriented UX patterns with real users before scaling.
Phase 2
Scale to Company-Wide Blue Badge
If pilot is successful, the framework replaces siloed tools like Embark and ASBX to standardize onboarding across all of Amazon.
Phase 3
Adapt framework to other plans
The framework can easily scale to support internal transfers, promotional roadmaps, and career transitions.
How might we unify fragmented platforms into a single framework that delivers a personalized, goal-oriented growth journey for every employee?
Deep-diving into the current UX
Once we aligned on the phased rollout scope, I audited the current onboarding journey. I needed to see exactly how the existing interface handled information and why it failed our users.
Problem 01
Lack of customization
A new UX designer and a developer received the exact same generalized corporate tasks, offering zero role-specific relevance.
Problem 02
Information overload
New hires faced a massive "wall of text" filled with generalized tasks, making it impossible to find role-specific information.
Problem 03
Static deadlines
Plans were locked into rigid dates (Day 1, Week 1) that didn't account for different role requirements or team-specific learning curves.
Improvement #1: modular learning chapters
To fix the rigidity problem, I needed to move away from a time-spent metric and toward an outcome-achieved framework. I led the shift to a modular architecture. This brings several benefits:
01
Outcome-Based Chapters
I replaced calendar weeks with Thematic Chapters (e.g., Understanding Culture, Understanding Your Team) to provide clear intent.
02
Scalable framework
By using "Goals" as the container, the architecture could now scale to any role, from corporate UX designers to operations leads.
03
Dynamic personalization
This shift allowed managers to filter out noise, ensuring employees only see content that directly contributes to their specific growth goal.
How does it work?
Instead of trying to redesign three massive software platforms all at once, our team made a strategic decision: design a single, standardized plan framework that could live inside any platform.
Improvement #2: The layout pivot
Transitioning to modular chapters introduced high data density. To find the right interface to present these tasks, I ran an A/B usability test that inverted my early design assumptions.
A
Variant A (The Sequential List):
A clean list view where clicking a task opened content in a new browser tab or window to keep the main list minimal.
B
Variant B (The Master-Detail Layout):
A split-pane dashboard. The left side featured a sticky milestone navigation track, and the right side instantly displayed the live content of the selected task.
Using AI to shape the research plan
Once the two directions were defined, I used Kiro to turn my early assumptions into a structured usability plan before recruiting participants.
01
From assumptions to testable questions
Kiro helped me challenge what I believed each layout solved and translate those assumptions into neutral research questions rather than leading prompts.
02
Coverage before sessions
I used it to pressure-test task scenarios, surface edge cases, and check that the plan could evaluate navigation, comprehension, and completion behavior across both variants.
03
Designer review remained essential
I edited the language, set the success criteria, selected the final tasks, and ran the study. AI accelerated planning; it did not determine what counted as evidence.
User research insights
While I initially assumed Version A would win due to lower visual density, users overwhelmingly preferred Version B (The Master-Detail Layout).
After the sessions, I used AI to cluster recurring behaviors and organize the findings into a stakeholder-ready research report. I reviewed every theme against the raw notes before turning it into a recommendation.
78%
Participants prefer varant B
Opening tasks in separate windows caused them to lose their spatial orientation, forcing constant back-and-forth clicking. The Master-Detail layout allows employees to browse and complete goals in a single workspace.
89&
Participants love the modular chapters
New hires faced a massive "wall of text" filled with generalized tasks, making it impossible to find role-specific information.
#1 Master-detail dashboard
I designed a unified split-pane interface. A left-hand navigation track keeps users within their broader growth journey, while the live execution pane on the right allows them to read, complete, and verify tasks within a single workspace.
#2 Modular growth chapter
Instead of an overwhelming wall of text, learning content is now grouped into modular, thematic chapters like "Understanding Your Team" or "Operational Deep-Dive."
#3 Personalized learning
Based on an employee's specific role, seniority level, and region, the system dynamically put together targeted tasks into their chapters.
Impact at enterprise scale
We launched a highly tracked soft launch to a pilot group of 46 new hires navigating 12 milestones. The goal-oriented framework radically shifted our historical 60% completion baseline.
01
Deep engagement
32 employees went above and beyond to complete over 90 specialized activities; 24 users pushed past 100 activities in record time.
02
Smashed the Baseline
The pilot group demonstrated significantly higher completion velocity and engagement depth than the legacy 60% benchmark.
Back-end framework design
To make this new experience work for the whole company, I designed a simple, 4-step setup wizard for managers and program leads. This tool allows them to easily bundle tasks, set due dates, and send out custom onboarding plans to specific roles or regions without needing help from engineers to rewrite code.
Ready for Global Scale
Although this backend tool is currently unlaunched, this additional systems work proves that our new framework is technically sustainable. It establishes the backend data architecture needed to seamlessly scale our goal-oriented ecosystem to all of Amazon's global workforce.
