Building Structure from Ambiguity at AWS

Building Structure from Ambiguity at AWS

How I created governance models and modular design components that unified experience across a fast-moving SaaS platform

Due to confidentiality, respect of privacy policies, and enterprise compliance, some visuals are recreated or simplified for portfolio purposes. A detailed version is available upon request.

CONTEXT

Role & Impact

UX Design Intern | 12 Weeks, May - Aug 2023 | San Francisco, CA

The Transformation: Reduced design inconsistencies, Unified team workflows, Accelerated dark mode transition, Improved UX-to-eng collaboration

THE PROBLEM

Fast-Moving Product Needed Structure Without Slowing Down

AWS was building a new enterprise low-code SaaS platform. As an early-stage product, the team prioritized rapid feature delivery to prove value and secure investment. Design consistency and governance took a backseat.

What I observed: Subdisciplines worked in parallel but weren't always aligned. This created fragmented experiences, redundant work, and growing design debt that would be hard to fix at scale.

No official design system governance. No end-to-end prototype showing how the experience came together.

The opportunity: Rather than seeing these as blockers, I saw a chance to contribute at a foundational level, creating alignment frameworks and laying groundwork for a scalable, systematic UX practice.

start query

Open SharePoint

across 14+ business libraries

Navigate to relevant documents

(if you know where they are)

Ctrl+F through hundreds of PDFs for key terms

hoping you're using the right search terms

Risk:

Missing critical safety information that could prevent incidents

complete query

Average Time:

30+ minutes per query

DISCOVERY

I Observed Design Crits and Had Coffee Chats to Understand Ways of Working

While onboarding, I had the unique opportunity to scope my own project direction. I carefully observed design critiques and set up coffee chats with colleagues to understand workflows.

What I discovered: Lack of shared structure was leading to fragmentation in both the design organization and the product itself.

THE SOLUTION

Three Foundational Contributions That Created Structure

Introduced a Scalable UX Governance Process

Designed Components with Tokens in Mind

Built a Clickable End-to-End Prototype

Collaborated with the design systems team to create new UI components across various states (rest, focus, pressed, error) with accessibility, dark mode, and semantic tokens in mind.

Impact: Contributed to a more token-driven approach that improves scalability and adaptability of the design system

Introduced a Scalable UX Governance Process

Designed Components with Tokens in Mind

Built a Clickable End-to-End Prototype

Collaborated with the design systems team to create new UI components across various states (rest, focus, pressed, error) with accessibility, dark mode, and semantic tokens in mind.

Impact: Contributed to a more token-driven approach that improves scalability and adaptability of the design system

ADDITIONAL WORK

Designed Retention-Focused Email Notifications

In the final weeks, I worked on customer-facing email notifications targeting low-code app builders to support user retention.

Design approach:

  • Surfaced critical app status info upfront

  • Used CTAs to guide users back into the product

  • Created positive emotional moments through visuals (confetti, personalized headlines)

Created several variations across traditional and more expressive visual styles, allowing stakeholders to choose what best fit the brand and user goals.

Design options:

Trust and Transparency

Uninterrupted Flow

Speed-to-Insight

Visible AI Reasoning

Users don't blindly trust AI, they trust what they understand. Transparency builds confidence, making AI feel less like a black box and more like a trusted partner working in the open.

How I Applied It:

  • Chain of thought interface showing which documents the AI searched and why

  • Source quality indicators helping engineers assess information reliability

  • Editable prompts allowing engineers to refine queries and retry

  • SQL query visibility for structured data requests

Visual Design Decision: Instead of hiding AI processing, I made it the focal point. Engineers could expand reasoning panels to audit every step of the AI's logic.

Trust and Transparency

Uninterrupted Flow

Speed-to-Insight

Visible AI Reasoning

Users don't blindly trust AI, they trust what they understand. Transparency builds confidence, making AI feel less like a black box and more like a trusted partner working in the open.

How I Applied It:

  • Chain of thought interface showing which documents the AI searched and why

  • Source quality indicators helping engineers assess information reliability

  • Editable prompts allowing engineers to refine queries and retry

  • SQL query visibility for structured data requests

Visual Design Decision: Instead of hiding AI processing, I made it the focal point. Engineers could expand reasoning panels to audit every step of the AI's logic.

01

02

03

04

05

Final iteration we landed on

06

07

08

09

LEARNINGS

How I Learned to Navigate Ambiguity and Create Structure

Bringing structure to fast-moving environments

I learned how to create alignment frameworks without slowing down progress. The key was making systems that enabled speed, not restricted it.

Thinking holistically while operating at detail level

This internship taught me to zoom in on component states and zoom out to see how design work flows across an entire organization.

Prototyping creates alignment, not just testing

The end-to-end prototype showed me that design artifacts can be powerful alignment tools. When everyone collaborates in the same file, shared understanding emerges naturally.

Initiative matters in ambiguous environments

Having the opportunity to define my own project taught me that the best contributions often come from identifying gaps others haven't addressed yet.

INTERNSHIP TAKEAWAYS

This internship taught me how to navigate ambiguity with optimism and initiative. I learned to collaborate across specialties, advocate for consistency, and contribute at a foundational level.

I'm deeply grateful to the team at AWS for trusting me with meaningful challenges and supporting my growth as a designer.