Product Design Leader
15+ years experience, 7+ years management�Specialities in B2B Saas, Marketplaces, AI-powered design
Portfolio
Yuwei Fu
Introduction
Results Driven
Collaborative
People Focused
Visionary & Strategic
Focus on outcomes—turning vision into measurable results through clarity and execution
Lead through partnership—aligning cross-functional teams and driving shared success
Grow people first—coaching, empowering, and fostering a culture of learning and trust
Inspire with a clear vision—shaping long-term direction and inspiring purposeful innovation
Inspire with vision, deliver with purpose, grow through people, and lead through collaboration.
My Leadership Philosophy
Oracle - UX Director �Driving AI Transformation in Enterprise Finance
Legacy Finance Systems
Fragmented Data
Outdated UX
AI Hesitancy
Inefficient Workflows
Context
Dispersed data prevents timely, strategic decisions
Complex, dated UI reduces user satisfaction and efficiency
Due to concerns over reliability, trust, and data governance
Manual, spreadsheet-heavy processes hinder agility
Defined Long-term AI Roadmap
Developed Oracle's AI adoption roadmap, defining long-term vision for enterprise.�Secured VP/SVP buy-in, aligning organizational objectives.
Leadership Action
Predictive AI
Gen AI
Agentic AI
Scaled Design Systems Efficiently �
Leadership Action
Led the Modernization Program applying the Redwood design system
8 apps
178 pages design delivered
400+ workflows
1.5 years
Built Team Domain Expertise
Leadership Action
Corporate Finance Learning
Oracle Application Training
User Research and Testing
Learning Progress Tracking
Drove Business Impact
Result
$4.7B �Revenue Growth
24% CAGR
97% �Most Willingness to Recommend
60% �Faster Finance Cycles
Drove FY25 revenue for Oracle's flagship ERP & EPM products through UX leadership.
Delivered AI-first user experiences that helped achieve top-tier customer satisfaction ratings from Gartner.
Shipped 10+ AI agents that automated core workflows, significantly reducing manual effort for customers.
Amazon - UX Manager �Driving Impact Across
Global Logistics,
Seller Experience &
Data Platforms
Complex Global
Selling Process
Context
new services, tools, programs causing confusion
calls seller support for onboarding
100+
50%
gave negative feedback on the compliance process
82%
chose small-parcel air freight causing high costs
75%
Influenced Executive Strategy
Leadership Action
improved experience, expanding logistics, and automating compliance
addressed friction points in seller journeys to earn seller trust
Global Seller Growth
3-year Plan
Seller Experience Outcome
Framework
consolidated 5 data apps into one and scaled
Business Data Platform
North Star Vision
Improved Organizational Efficiency
Leadership Action
Scaled Figma Adoption
45+ designers onboard
Standardized design process
7+ teams
Through a guide for design collaboration, handoff, experimentation, and instrumentation.
Leadership Action
Built Design Community
�Organized design events
Avg. 160 attendees
Improved Organizational Efficiency
Scaled High-Performing Teams
Empowered designers and led with data-driven results.
Leadership Action
Team Building Excellence
10+ designers
across various initiatives, leading to 3 promotions
Leadership Recognition
exceeded Amazon's average Manager Satisfaction scores
above benchmark
11%
Contributed Impact at Immense Scale
Result
$300B+�Gross Merchandise Value
41.7M �Error Reduction
113% �YoY Growth
Launched 80+ features enabling 210K+ sellers to fulfill 2B+ units annually.
Team effort significantly reduced seller-caused inventory errors.
UX vision for Global Logistics increased inbound units and cut shipping costs by 38%.
Appendix
Oracle Enterprise SaaS AI & Finance Transformation��Amazon Global Logistics
Oracle Touchless Expenses Mobile App
Case Studies
Oracle Enterprise SaaS�AI & Finance Transformation
Shift finance toward touchless operations, predictive insights, and connected actions.
Case Study #1
AI revolution in Finance Workflow
Task
Results
Actions
Situation
Context
Reimagined finance workflow
Designed AI adoption model (predictive AI, GenAI, agentic AI)
Embed predictive, generative, and agentic AI into core finance workflows and enabling CFOs to focus on strategy.
Led UX for AI agent integration�Drove GenAI adoption �Partnered with design system team
Manual processes, fragmented data, and regulatory demands hinder agility. ��AI is a revolutionary force.
Context
Through aligning data, workflows, and design patterns into a seamless experience that empowers finance teams to act on insight, not input.
ERP & EPM Integration
Purchase Order
Order Fulfillment
Shipment
Invoice
Payment
Accounting & Reconciliation
Planning, Analysis & Report
ERP→”What”
EPM→”Why”, “What If”
Daily operation: Supply Chain, Sales & Marketing, Finance, HR, IT
Strategic Management, Forecasting
Goal Setting
Context
Shared goals/
design principles
Trust
Efficiency
Adoption
Business Goals:
Transform finance into strategic partner
Automate workflows end-to-end
Differentiate Oracle ERP/EPM with AI leadership
Build customer trust in AI (explainable, transparent)
UX Design Goals:
Embed AI seamlessly into existing finance workflows.
Design for trust with transparency, oversight, and explainability.
Ensure consistency & accessibility via Redwood design principles.
Support progressive adoption
Shared Goals:
Trust
Efficiency
Adoption
Before vs After
Implemented a standardized design and innovation approach.
Innovation Cycle
Leadership Action
Predictive AI
Solution
AI Transformation
Results
Auto predict seeded forecasts, Predictive cash flow models, Trend & anomaly detection
Up to 70% shorter planning cycles
+12% forecast accuracy improvement
Builds trust in AI with data-driven forecasts and measurable accuracy gains
Machine Learning Explanation
Predictive cash flow models: ML, Driver, Trend
Generative AI
Solution
AI Transformation
Results
Automated narrative drafts, Variance explanations, AI-generated visuals & charts
Reduced reporting effort for QBRs,�Expands FP&A value from reporting numbers to communicating insights
Automated narrative drafts
AI-generated visuals & charts
Agentic AI
Solution
AI Transformation
Results
Advanced prediction with external drivers,�Agent orchestration (forecast → journals → reporting)
Continuous monitoring of variances,�Touchless analysis & workflows,�Unlocks end-to-end automation
Advanced prediction with external drivers
Contextual data exploration for scenarios
Finance workflow AI transformation lists
Workflow
Problem
Solution
Manual accruals and journal entries slowed quarter-end close.
Accounting & Reconciliation
Reconciliation and transaction matching consumed weeks of effort.
Planning and forecasting relied on spreadsheets, often inaccurate and disconnected from real-time drivers.
Planning, Analysis & Reporting
Gen AI Predictive Insights Demo Video
Gen AI Narrative Reporting Demo Video
Quarterly reporting required weeks of manual narrative creation.
The Ledger Agent spotted anomalies and, through natural language prompts, auto-generated journal entries, eliminating spreadsheet-based manual work.
AI Predicted Matching & Attributes automated high-volume reconciliations and assigned risk attributes, cutting reconciliation cycles by up to 50%.
The Advanced Prediction Agent used multivariate models with financial, operational, and external drivers to predict revenue and cash flow.
GenAI for Narrative Reporting drafted QBRs and management reports, producing clear text, causal analysis, and dynamic visuals. CFOs could edit instead of starting from scratch.
Employees spent hours submitting receipts, and finance teams wasted time auditing and correcting non-compliant expenses.
Integrated AI Expense Policy Agent to auto-capture corporate card transactions, match them to receipts, and pre-validate against company policies.
Manual requisition, invoice processing, and payments were slow and error-prone.
Procure-to-Pay
The Document IO Agent automated requisition creation from vendor quotes, processed invoices in multiple formats, proposed corrections, and triggered payments with early discount detection.
Amazon Global Selling�Global Logistics
Allow sellers to ship from Asia directly to fulfillment centers in the US, UK, EU, & Japan
Case Study #2
Major sellers are based in Asia
User Types
Goal
Problem
Amazon Marketplace
Context
Sellers, Business Owners,�Logistics Coordinators, Operators
Drive Marketplace growth by enabling sellers to ship direct-to-fulfillment centers worldwide.
75% of shipments relied on expensive air freight, causing limited selection and high cost of goods.
57% of sellers earning $1M+ annually are based in China.
India (~7%) & Japan (~6%) account for ~13% of all global sellers.
Challenges
Global logistics
Situation
Sellers relied on brokers with little control or visibility.
Shipment creation workflows were repetitive and error-prone.
Customs clearance slowed down global expansion.�
Onsite research physical steps of global shipping
UX Manager
My responsibilities
Task
Built and led multi-region team (Seattle, Beijing, Shanghai)
Led design directors of end-to-end UX for Global Logistics platforms:
�
Seller Central
Shipper Central
Operator Console
Seller Central
0-1 enabled global shipment workflow iterations
User Types:
2018 Aug
Launched a tech-based solution
2018 Nov
Experience enhancement
through research
2019 Jan
Experience enhancement
through research
2019 May
Enabled CN payment & compliance setup localization for CN sellers
2019 Jul
First-time & repeated experience
Discussed booking systems parity btw Seller & Shipper Centrals
Shipper Central
User Types:
Operator Console
User Type:
Seller Segmentation
Research
To define project experiences, it’s critical to define more nuanced seller segmentation in terms of company size, structure, and their roles.
Journey Map
Research
50+ seller interviews
Quarterly surveys
�Journey maps covered�7 roles (sellers, coordinator, operators, etc.) & �3 platforms based on the research
�
Business Roles
Research
Defined Business Roles across different Selling Partner types
�
User Testing
Research
Roundtables at Amazon Global Seller Summit in Shanghai
�
Templatized Booking Preference & Packing
Design
28% time saved
70% manual input eliminated
Shipment Dashboards
�Real-time visibility on cost,
timing, and performance.
Design
Compliance Portal
All compliance-related tasks across product compliance domains.
Design
3-year roadmap
Partnered with Ops, Tech, and Sales for a 3-year logistics UX roadmap
Enabled sellers shipping from Asia directly to fulfillment centers in the US, UK, EU, & Japan.
Collaboration
Global teams across US, CN, IND
Traveled to China for the sellering event, and collected CN logistics providers info for research. Met with cross-functional teams (China & India) for UX roadmap alignments.
Impact
Enabled Amazon’s most important seller regions to thrive globally.
Conclusion
210K+ �Sellers fulfilled 2B+ units
38% �YoY reduction in �operating costs
4.5/5�Seller satisfaction �improved from 3.5
Learnings
Conclusion
Lack of seller segmentation: introduced role-based journeys.
Data gaps: designed new performance metrics (time-on-task, completion, satisfaction).
Cross-regional complexity: established UX mechanisms through
I defined metrics to measure seller behaviors and reduce 28% time on tasks through a booking preference template. The metrics include click streams, time on tasks, and completion rates.
Amazon Data Technologies�Business Data Platform
Accelerating data-driven business, enabling the next generation of analytics and machine learning technologies at scale
Case Study #3
Challenges
Engineering-driven products
Context
UX not involved initially, poor usability.
Scattered workflows
Users juggled 5+ apps, losing time.
Low productivity
Tasks took up to 2 weeks, often required manual support.
Before
Goals
Business Goals
Context
UX Design Goals
Accelerate Amazon’s data-driven decisions.
Reduce time-on-task and reliance on manual support.
Strengthen governance + security for data at scale.
Simplify data discovery and management.
Provide centralized workflows (search, notifications, table creation).
Establish UX process + culture in an engineering-first org.
Team and Practice Leadership
Established UX Practices
Defined UX Design Process
Cross-Functional Influence
Built UX Team
Task
UX Office Hours, Voice of User, Bar Raiser, stakeholder reviews.
Unitized the iterative process to understand the user, challenge existing assumptions, and redefine problems.
Represented UX in annual operating planning (VP-level). Created shared team goals of establishing UX framework.
Recruited, mentored, and scaled first design org in Business Data Technologies in one year.
User Personal
Data Producers: publish/manage datasets, often overwhelmed by metadata chaos.
Data Consumers: �search + join datasets for analytics/ML.
Data Engineering Managers: monitor workflows, resolve breaches, ensure compliance.
Research
Data Producer Journey Map
50+ studies, mapped producer/consumer workflows:�
Research
Universal Search
Design
One-stop search across datasets to cut discovery from days to minutes.
Notifications
Design
Centralized communication + alerts to reduce ticket chaos (1,000+ avoided).
Table Creation Workflow
Design
Simplified schema + table creation from tasks completed in <15 mins vs. 3–7 days.
Impact
Consolidated 5 apps into 1 platform
Result
1.5K �Petabytes of Data Catalog
1 petabyte = 1K terabytes, �FB generates 4 petabytes of data per day
62K �Monthly Active Users
15 mins �Time-on-task Reduction from 5 Days
309K Days Saved/month
Oracle ERP�Touchless Expenses
Mobile APP
Highly automated process for managing corporate expenses
Case Study #4
Corporate Expenses Challenge
Goals
Approvers & Auditors
Spenders
Context
Achieve touchless operations through advanced automation
Ensure users get expenses reimbursed from avg. 9 days to 1 day through corporate card integration
Lack of visibility
Rick of fraud
Poor budget control
Manual and time-consuming tasks
Lost receipts and forgotten expenses
Delayed reimbursements
Vanguard project for touchless
5 rounds of internal pilots from 26 to 723 JPM cardholders
Task
Delivered a nearly touchless & automated expense workflow for Spenders
26K
Expense items auto-submitted �(75% of total)
77%
Overall expense �matching rate
(Taxi/Rideshare: 100%, Meals: 75%, Accommodation: 40%)
80%
User satisfaction �based on surveys
Improvements
Action
Combine Expenses landing page with Search page features
Bring “Add Expense” to the landing page from the bottom navigation
Improve bottom nav spacing and surface other key pages
Before
After
Search page
Expenses Landing
New Expenses Landing
Policy Agent
Action
Provide out-of-policy reasoning tailored to the user’s role and current location
Offer a more user-friendly alternative to scanning FAQs and policy documents
Expense Policy Agent
Expense item page
Expenses Agent
Action
Enable batch receipt upload through email forwarding
Improve receipt matching to 95%+ accuracy (from 77%) including foreign languages, handwriting, and tips
Email batch receipt photos
Email a receipt photo with handwriting
PWA Barrier
Learning
PWA (mobile) usage is less than 20% due to burdensome authentication.
To enhance touchless experience, leadership decided to expand email covering more use cases instead of PWA.
Email batch receipt photos
PWA Authentication
Complete Expense Workflow
Next Steps
Expend workflow to Approve, Audit, and Pay
Apply Expenses and Policy Agents across workflow
Expense Audit workflow
Team’s weekly expenses email for Approvers (managers)
Receipt Creation Flows
Examples
Google AdWords On-boarding Learning Game
McDonald’s ordering system
JP Morgan Chase web application
Nickelodeon app
FuboTV streaming web application
Agency
Adwords Onboarding Tutorial
Google AdWords marketing team wanted to design a standalone interactive AdWords on-boarding web app.
Help un-tech savvy small business owners, who plan to get started in online advertising but feel AdWords is hard, confusing and expensive.
The goal was not only to usher them through the campaign creation funnel blindly, but they came away with true concept understanding and were able to complete an effective campaign.
User Research
Action
We talked to frustrated churned small business owners who have struggled with AdWords and stopped using the product. These SMB owners are extremely time-constrained. Research shows that SMBs spend approximately 2 ½ hours/week managing their company’s marketing.
Design Solution
Action
I came out with an AdWords City to simulate both Google search interface with online browsing behaviors and real world small business development.
AdWords City allows small business owners using AdWords to attract customers to their business. The interface would be able to demonstrate how customers come into your store from searching for you on Google.
User Testing
Action
We had a chance to conduct user testing with onsite business owners in Kansas City as part of a larger Google business outreach expo.
At the end of the sessions, we talked seven individual SMBs from age 25-60 and work at various industries, resulting in positive feedback and understanding how to use the AdWords platform.
Kiosks Ordering System
20%+ Customer spends on McDonald's kiosk
Phone
Thank You