In-depth case studies from my UX engineering work at Coursera, demonstrating user-centered design processes, research methodologies, and measurable outcomes.
Improving Course Accessibility for Screen Reader Users
Coursera · 2023 · 6-month project
Problem Statement
Screen reader users reported significant barriers navigating course content on Coursera. Interactive quiz elements, inaccessible math notation, and unlabeled graphics created major obstacles. Internal audits revealed only 62% WCAG 2.1 AA compliance across core learning pathways.
Research & Discovery
Conducted mixed-methods research with 24 screen reader users over 8 weeks:
Moderated usability testing sessions with NVDA, JAWS, and VoiceOver users
Analysis of 1,200+ support tickets related to accessibility barriers
Automated accessibility scans across 5,000+ course pages using axe-core
Competitive analysis of accessibility practices at edX, Udacity, and Khan Academy
Figure 1: Redesigned quiz interface wireframe with semantic structure and visible focus indicators
Figure 2: Accessible math notation prototype with MathML + screen reader integration
Solution & Implementation
Redesigned quiz components with proper ARIA roles, live regions, and keyboard-only navigation
Implemented MathML with MathJax fallback for accessible math notation rendering
Added automated alt-text generation pipeline for course images using AI + human review
Built custom skip navigation and landmark regions improving screen reader efficiency
Created comprehensive accessibility testing guide for course authors
Outcomes & Impact
WCAG 2.2 AA compliance increased from 62% to 96% across core learning pathways
Screen reader user satisfaction scores improved by 42% (post-launch survey, n=180)
Accessibility-related support tickets decreased by 55% within 3 months
Adopted as company-wide accessibility benchmark for future product development
Redesigning the Learner Dashboard for Engagement & Inclusion
Coursera · 2022 · 4-month project
Problem Statement
Learner dashboard engagement was declining, with only 34% of users returning weekly. Users felt overwhelmed by information density and frustrated by unclear progress indicators. The dashboard had significant accessibility gaps for users with cognitive disabilities and screen reader users.
Research & Discovery
Conducted 18 semi-structured interviews with learners across diverse backgrounds and abilities
Analyzed behavioral data from 50,000 learners to identify drop-off points and engagement patterns
Performed cognitive walkthroughs with users with ADHD, dyslexia, and anxiety disorders
Heuristic evaluation against Nielsen's 10 usability heuristics and WCAG 2.2 cognitive accessibility guidelines
Figure 3: Dashboard wireframe with simplified card layout and clear progress indicators
Figure 4: High-fidelity prototype with warm, welcoming visual design and personalized content
Solution & Implementation
Simplified information architecture with progressive disclosure, reducing cognitive load
Introduced personalized greeting and adaptive content prioritization based on learner behavior
Redesigned progress indicators with clear, scannable visual cues and text alternatives
Implemented "Focus Mode" reducing visual clutter and distractions for neurodiverse users
Ensured full keyboard navigation and screen reader compatibility with semantic HTML
Outcomes & Impact
Weekly active user retention improved from 34% to 62% within 8 weeks of launch
Task completion rate for key actions increased by 45%
Accessibility satisfaction score improved from 3.2/5 to 4.6/5
"Focus Mode" featured as accessibility best practice in Coursera's design guidelines
Building an Accessible Design System at Scale
Coursera · 2021 – 2024 · Ongoing initiative
Problem Statement
With 40+ product teams shipping features independently, Coursera's UI had become inconsistent. Components were rebuilt from scratch across teams, leading to fragmented accessibility, visual inconsistency, and significant technical debt. Designers and engineers lacked a shared source of truth.
Research & Approach
Audited existing UI implementations across 12 major product areas, identifying 200+ unique component variations
Conducted a design system maturity assessment using the Design System Maturity Model
Interviewed 24 engineers and 12 designers to understand pain points and workflow needs
Established accessibility as a foundational requirement with defined acceptance criteria for each component
Figure 5: Component documentation in Storybook with accessibility guidelines and live examples