--.--.--
Case File — Accessed
Applied Materials
Record 01 / 04
Justin Lee — Applied Materials · 2020 – 2025
Record Open
--:--:-- Local
Case 01 / 04
Case Study

Applied
Materials

Role
UX Designer / Product Manager
Scope
Enterprise HR · 26 Countries
Timeline
2020 – 2025

Applied Materials HR portal
HR Portal & AI Journey

Applied Materials needed to redesign its fragmented HR experience for 36,000+ employees across 26 countries. Employees struggled with disconnected portals, inconsistent navigation, and outdated content, leading to heavy reliance on HR support. The project evolved from a portal redesign into a broader initiative focused on self-service and AI-powered employee support.

As UX Lead and later Product Owner, I led research, UX strategy, interaction design, and AI initiatives across the platform. I designed scalable self-service workflows, predictive-intelligence support tools, and helped launch an AI assistant in Microsoft Teams, improving search adoption from 3% to 95% while reducing HR dependency at scale. This project spanned 5 years and included 2 major portal design updates as well as an AI implementation.

96%Search Utilization
34%Reduced Tickets
+43NPS Growth
28%Time to Resolution

From the beginning of the project in 2020 I realized that the optimal user experience was to have one place for all employee digital experiences. What we were facing in HR was the same across all departments. To combat this, I coordinated with designers from IT to align our designs and started designing for the future — a single unified portal experience.

AiMEE — Gen AI implementation
4.4★Star Rating
88%Output Accuracy
523User Upvotes
7minNavigation Saved

The Unified Portal is currently going through executive leadership buy-in. We want to continue to expand the platform to all departments and create a single digital experience for all Applied Materials employees globally.

Future vision, under approval process
Deeper Dive
Phase 1  ·  2020 – 2022 Consolidation & Clarity
Context

25 portals. One company. No shared language.

30,000+ employees globally. Internal tools across HR, IT, Legal and Facilities all on ServiceNow — but spread across 25+ country portals with no unified navigation. Search utilization sat at 5%.

25+Country portals
5%Search utilization at start
2–3hrHR lost weekly to tier 0/1 questions
Before — original portal After — Iteration 1
⟵⟶
Before ← drag → Iteration 1
Research

Six methods. Four root causes.

Interviews, focus groups, surveys, ticket analysis, usage analytics, and A/B testing converged on the same problems: search wasn't trusted, navigation caused abandonment, and self-service couldn't scale culturally.

Affinity mapping — discovery synthesis
Trade-offs

Consolidate first. Personalize later.

We deliberately cut localized COE landing pages and per-persona features to stay focused on foundational consolidation — the work that would make everything else possible.

Focused on
  • Cross-site consolidation
  • Unified branding
  • Shared services modernization
Cut back
  • Localized COE pages
  • Per-persona personalization
  • SEO optimization
Iteration 1

One front door for HR.

All HR functions — Onboarding, Employee Engagement, Case Submission — under a single homepage with a Campaign Carousel, category sidebar, and unified search.

Iteration 1 — New homepage
Iteration 2

Personalization and COE depth.

Unified taxonomy, personalized shortcuts, and dedicated COE pages. Smart Form consolidated 50 forms into one funnel — cutting routing time by 5 hours and reducing misdirected tickets by 72%.

Guided Self Service
Smart Form
Impact

26 portals → 1 platform.

26Portals unified
31%Search CTR increase
15%Fewer HR escalations
28%Faster ticket resolution
Phase 2  ·  2023 – 2024 AI Implementation
Why AI

Limits a portal couldn't solve.

Content required interpretation. HR bandwidth couldn't scale. Users needed to be met where they already were. Technology: ChatGPT + ServiceNow, delivered inside Microsoft Teams.

01 Content Requires Interpretation
02 HR Bandwidth Constraints
03 Meet Users Where They Are
Decisions

Simplified IA over surface-level shortcuts.

Quick fixes would have compounded long-term complexity. We chose to simplify the information architecture and build governance frameworks — slower short-term, the right foundation for AI.

Strengths
  • Strong foundational IA
  • Clear user behavior data
  • Executive buy-in for AI
Weaknesses
  • Limited team capacity
  • Knowledge base not AI-ready
  • Change fatigue risk
Opportunities
  • Embed AI into Teams
  • Governance frameworks
  • AI-ready knowledge base
Threats
  • AI hype vs. readiness gap
  • Organizational stability
  • Vendor dependencies
AiMEE

An HR agent inside Microsoft Teams.

Deployed where employees already worked — answering HR questions in real time, without a portal visit. Designed around action-oriented outputs, measurable accuracy, and a trust-building metrics framework.

AiMEE — Teams conversation interface
Metrics Framework

Defining what trust looks like in numbers.

The hardest part wasn't building the AI — it was knowing whether it was working. I designed a four-category framework to measure trust, usage, sentiment, and content quality across every interaction, then mapped every signal to a specific metric in the dashboard.

01
Trust Building
Star ratings, thumbs up/down ratio, rating distribution by question type and department
02
Usage
Active users, sessions per week, questions per session, return rate and adoption curve
03
Sentiment
Open-ended feedback synthesis, qualitative theme clustering, AI aversion signals
04
Content Quality
Knowledge base coverage gaps, broken reference links, content freshness and accuracy scores
Metrics mapping — four-category tracking framework
Impact

4.43 stars. 88% clarity.

4.43★Average user rating
88%Output clear and actionable
523Upvotes vs 233 downvotes
What worked
  • Quick access right in Teams
  • Answers were clear and helpful
What needed work
  • Not all reference links working
  • No historical context learning
Phase 3  ·  In Progress Enterprise Unification
The Problem

HR wasn't the only silo.

After five years of improving the HR portal, the same fragmentation appeared across IT, Facilities, Legal, Procurement, Travel, and Business Units. HR was still the catchall. The insight that had been forming since 2020 became the vision for Phase 3.

Vision

One door, many systems.

The Applied Materials Employee Center unifies HR, IT, Facilities, Legal, and Procurement under a single portal — addressing the root cause rather than optimizing individual silos. Currently in executive review.

Reflections What Five Years Taught Me
Conclusion
01 Continued iterations may lead to an entirely different solution than the one you started with.
02 Full product ownership — from road mapping to execution — compounds over time.
03 It takes good data to train an LLM. Trust is the key factor. How you build it matters more than the model.
04 Missed AI expectations cause aversion. Over-reliance without expertise is the opposite risk.
05 Internal teams sometimes understand the problem better than outside consultants. Five years of context is a competitive advantage.