AWS Learning Assistant Case Study

Guided Learning Within AWS Skill Builder

Learning Assistant hero mock

My Role

  • Senior/Sole UX Designer
  • User Research & Usability Testing
  • Accessibility Notation (a11y)
  • Stakeholder Collaboration

Working Group

  • Product Managers
  • Principal Engineers & Front-End
  • Applied Scientists / LLM Engineers
  • Accessibility Specialists

Stakeholders

  • Executive Leadership
  • Design Team
  • Engineering Team

Customer Jobs to Be Done

Understanding what users are trying to accomplish and the outcomes they seek

Ask questions instantly

Unblock progress during labs and digital classrooms without breaking focus.

Deepen understanding

Get clear explanations of AWS services, trade-offs, and real-world applications.

Receive hints and explanations

Guidance for code, troubleshooting, and complex concepts in context.

Learn without judgement

Ask anything in a private, supportive space free from peer pressure.

Challenges

Lack of Real-Time Support

Learners in digital classrooms and labs didn’t have immediate help when stuck, unlike instructor-led training with live Q&A.

Knowledge Retention

Training often felt one-way and transactional, creating gaps in reinforcement and confidence.

Scattered Resources

Learners had to leave their training flow to search for answers, disrupting focus and progress.

Challenges illustration

Insight & Strategy

We envisioned the assistant as a personal study coach: approachable, clear, and context-aware. Three strategies guided design:

Awareness & Onboarding

Surface value early with contextual prompts and zero-state guidance.

Iterative Refinement

Continuously tune tone, detail, and UI based on learner feedback.

Accessibility & Inclusivity

Ensure usability with screen readers, keyboard nav, and WCAG guidelines.

Execution

Design Leadership

Drove end-to-end UX, from wireframes through high-fidelity prototypes, aligning daily with PMs and engineers.

Onboarding

Created a simplified first-use experience that introduced prompts based on training context.

Feedback Loops

Added thumbs up/down plus predefined categories and β€œOther” for more nuanced learner input.

Interface Design

Delivered a clean, unobtrusive chat widget integrated into labs and classrooms with contextual awareness.

Accessibility Advocacy

Marked up designs for a11y, ensuring compatibility with screen readers and inclusive patterns.

Designs

Design screenshot 1
Design screenshot 2

Example Scenario

Sue learns with real-time support

  • While completing a lab, Sue highlights a line of code and asks the assistant what it does. She gets a clear explanation with links to AWS docs.
  • During a course, she asks for a hint on a tough question and receives guidance without breaking her study flow.
  • When reviewing concepts, Sue uses the assistant to reinforce tricky topics with simplified explanations.
  • At the end of the week, she sees how the assistant’s support helped her complete more labs with confidence.

Results & Impact

CSAT
4.7/5
Learners praised clarity and supportive tone.
Engagement
+38%
Goals, reminders, and streaks strongly supported learning.
Completion Rates
+25%
Higher lab and course completions with assistant.
Feature Adoption
78%
Learners used assistant during their training sessions.
Previous
Previous

Skill Builder Mobile App

Next
Next

AWS Jam