Case Study
Approach
- Facilitated internal and external workshops with 22 participants, generating 251 opportunity ideas
- Conducted AI-assisted competitive research across 4 leading e-commerce platforms and 4 postal services
- Ran viability and impact assessments across 6 opportunity areas, narrowing to 17 high-value concepts
- Created concept cards to consolidate 17 ideas into 4 prioritized GenAI initiatives
- Illustrated 4 service blueprints to map end-to-end customer and operational impact
- Partnered with data scientists to define conceptual architectures and data dependencies
- Delivered executive readouts and storytelling assets to align stakeholders and drive decisions
Outcomes
Delivered a clear, actionable roadmap grounded in customer needs and technical feasibility, aligning product, design, and engineering stakeholders around a scalable, future-ready AI vision for USPS digital platforms.
Defined a Future-Ready AI Strategy
Led a comprehensive GenAI/AI opportunity assessment, identifying 4 priority AI focus areas aligned to USPS business goals and customer experience improvements.
Accelerated Executive Buy-In
Created high-impact storytelling assets, including journey narratives, service blueprints, and concept cards that secured cross-functional alignment across product, technology, and leadership.
Brought Human-Centered Design to AI Strategy
Applied HCD methods to ensure GenAI concepts addressed real user needs and meaningful friction points across USPS’s digital ecosystem.
Improved CX & Operational Efficiency Forecasts
Identified AI use cases with measurable upside, including reduced customer service load, improved delivery transparency, higher postal store conversion, and increased self-service success rates.
Client
United States Postal Service
Firm
AFS
Year
2025
Role
Design Director
Scope
Led a small, cross-functional team and partnered closely with product, engineering, and data science stakeholders.
Facilitated internal and external workshops, guided research and synthesis, and drove strategic alignment from discovery through executive readouts
Visuals
