Product development workflow configuration and execution

When building Amazon's PLM tool for Private Brands, the platform needed to support robust workflows with over 250 steps to manage and track product development work across three major organizations—Softlines, Hardlines, and Consumables. Each organization had different processes, personas, and definitions of "product," yet they all needed to execute complex, constantly changing workflows.

The core challenges included:

• Lack of Visibility: Business requirements and processes changed frequently, but there was no visibility into where bottlenecks existed in the current process
• Compounding Effects: Without visibility, business leaders couldn't accurately determine capacity, resource plan to manage bottlenecks, or understand which bottlenecks had the most significant impact
• Process Fragmentation: Three different organizations had different processes for the same activities, making standardization nearly impossible
• Scale & Complexity: A single product (e.g., a shirt) could spawn 2,000 child products (10 styles × 20 colors × 10 sizes), each requiring distinct workflow management
• No Metrics: The business operated on email, Excel, and institutional memory with no data-driven insights into process efficiency

As the Principal UX Designer, I was responsible for designing the end-to-end workflow system from scratch, including both the self-service configuration tools for business admins and the execution experience for end users. My responsibilities extended far beyond typical design work:

Design & Architecture:

• Designed every aspect of the workflow engine including BPMN-informed configuration tooling
• Created self-service workflow design tools enabling business admins to configure complex workflows without developer support
• Designed task execution experiences for internal and external users (vendors) with conversation threads and collaboration features
• Established design patterns, contributed to the pattern library, and created Figma component libraries

Technical Learning & Contribution:

• Learned BPMN (Business Process Model and Notation) standards and Camunda workflow engine to design within technical constraints
• Pulled the code repository, built React components, and shipped production features when critical usability issues weren't roadmapped
• Collaborated with engineers to ensure implementation matched design vision while remaining technically feasible

Strategic Leadership:

• Navigated complex organizational politics across three business organizations with competing priorities
• Organized "walk the store" sessions with senior leadership that led to major roadmap pivots toward UX improvements
• Influenced roadmap decisions, determined scope and tradeoffs, and advocated for long-term UX vision over multiple years
• Served as cross-org connector with holistic knowledge of all features, helping PMs leverage existing patterns to launch faster

I designed a comprehensive, self-service workflow system that gave Amazon Private Brands the ability to configure, execute, and measure complex product development processes without engineering support. The system was built on BPMN standards and consisted of two major components:

1. Self-Service Configuration Platform

Business admins could design workflows visually using BPMN-informed flow diagrams, configure robust task definitions with assignment rules, conditional logic, and dynamic expansion for complex scenarios. The system supported shareable components, allowing teams to build libraries of reusable workflow patterns.

2. Execution & Collaboration Experience

End users received clear visibility into what work needed to be done, with tools and resources to complete tasks efficiently. The system included conversation threads (both internal and external), task assignment across a dozen contributors, and deep linking to relevant data and forms.

The Strategic Tradeoff

The initial MVP required hard tradeoffs: no batch editing, no contextual task execution, and users were deep-linked to existing pages that weren't optimized for task flow. While painful in the short term, this compromise gave the business immediate visibility into task-level dwell time and bottleneck identification—data that didn't previously exist. This visibility unlocked buy-in for deeper UX investments and enabled data-driven process optimization.

The workflow system transformed the business from speculative understanding to measurable, data-driven decision-making, providing clarity on capacity planning, process inefficiencies, and resource allocation.

The task execution experience was designed to help users understand what work needed to be done and provide efficient access to the tools and data required to complete it. Users could see all assigned tasks, filter by status or due date, and access detailed task information.

Key Features:

• Task Dashboard: Clear overview of all assigned tasks with status, due dates, and priority
• Task Details: Instructions, resources, conversation threads, and completion requirements
• Deep Linking: Direct access to relevant forms and data views (initial MVP approach)
• Collaboration: Internal and external conversation threads to resolve blockers
• Status Tracking: Real-time visibility into task progress and dwell time
• Notifications: Email and in-app notifications for new assignments and updates

Evolution: The initial MVP used deep linking to existing pages. Later iterations added contextual editing, batch operations, and the ability to complete tasks without leaving the task view—significantly reducing cognitive load and improving efficiency.

To improve efficiency and standardization across workflows, I designed a component library system that allowed admins to create reusable workflow patterns. These shareable components could be configured once and reused across multiple workflows, reducing configuration time and ensuring consistency.

Benefits of shareable components:

• Consistency: Standardized patterns across different product families
• Efficiency: Configure once, use many times
• Maintenance: Update a component and all workflows using it inherit the changes
• Best Practices: Capture and share proven workflow patterns across organizations

Task configuration was the most critical and complex part of the workflow system. Each task needed to support a wide variety of scenarios across different product types and organizational requirements.

Configuration options included:

• Assignment Rules: Define who receives the task based on role, product type, or dynamic conditions
• Dynamic Expansion: Automatically create tasks for complex entities (e.g., one task per color variation)
• Conditional Logic: Show or hide tasks based on product attributes or previous task outcomes
• Instructions & Resources: Rich text editor for task instructions, attached documents, and helpful links
• Due Dates & SLAs: Configurable timing rules with escalation paths
• Data Scope: Define which entity fields the task allows users to edit
• Validation Rules: Required fields and data quality checks before task completion

The configuration interface needed to handle hundreds of data fields across many different entities (products, materials, certifications, claims, catalog data, etc.), often with interdependencies or relationships between entities.

Data-Driven Decision Making:

• First-time visibility into task-level dwell time and bottleneck identification
• Enabled capacity planning and resource allocation based on actual metrics
• Clear understanding of which process inefficiencies had the most significant impact

Operational Efficiency:

• Self-service configuration eliminated the need for engineering resources to create or modify workflows
• Workflows could be adjusted quickly in response to changing business requirements
• Standardized processes across three major organizations while maintaining necessary flexibility

Strategic Outcomes:

• Provided the foundation for process optimization and automation initiatives
• Consolidated fragmented email and Excel-based processes into a single source of truth
• Enabled compliance tracking and audit trails for regulatory requirements
• Created scalable infrastructure supporting billions of dollars in product development

Long-Term UX Vision Adoption:

After years of advocacy, the "walk the store" sessions with senior leadership led to immediate roadmap pivots. Key UX improvements were finally prioritized, including AG Grid table implementations, batch actions, data inheritance, and contextual editing capabilities—features that dramatically improved usability and reduced user attrition.