Project details

Role: Product Equity Designer (Design Lead)

Team: PMs, Engineers, UXR, Content Design, Data Science, Legal, Operations, Support, Policy, and NGOs

Timeline: Sep 2022 – current

Impact: Improved reliability and trust in critical moments of the ride experience, reducing friction during pickup and increasing successful trip completion for riders with accessibility needs

Scope: End-to-end system design — from problem framing and research to defining scalable patterns, validating solutions, and supporting implementation across multiple markets

Collaborations: Partnered cross-functionally to align user needs, marketplace dynamics, and regulatory constraints — integrating accessibility signals into core matching, pickup, and feedback systems to ensure consistency and scalability

・・・・・・・・・・・

Executive summary

Accessibility work often starts as isolated fixes — a new flow here, a content change there. But fragmented solutions don’t scale across markets, teams, and regulatory environments.

I led the transformation of Uber’s accessibility experiences from reactive patches into structured, reusable systems. This work standardized how riders disclose accessibility needs, operationalized pickup support, and created measurement loops that allowed teams to improve reliability at scale.

Instead of shipping features, I built foundations.

・・・・・・・・・・・

The Problem:

Fragmented Accessibility at Scale

Different users handled service animal disclosure differently. Blind and low-vision riders relied on ad-hoc communication at pickup. Drivers lacked structured context. Support teams worked from proxy metrics instead of experience-level signals.

The result:

• Inconsistent rider experiences

• Increased pickup failures

• Regulatory risk across markets

• Reactive instead of proactive design decisions

The challenge wasn’t to design one better feature.

It was to design a scalable system that could:

• Work across multiple countries

• Respect marketplace dynamics

• Balance rider agency and driver context

• Adapt to regulatory differences

• Improve measurable reliability

・・・・・・・・・・・

The System Strategy

I reframed accessibility from a collection of flows into a layered system composed of:

1. Structured Disclosure

2. Context-Aware Pickup Support

3. Measurement & Feedback Infrastructure

4. Marketplace-Safe Matching Signals

Each initiative below represents a system module — not a standalone feature.

・・・・・・・・・・・

Module 1: Standardizing Service Animal Disclosure Across Markets

Service animal experiences are high-stakes. A small UX inconsistency can lead to denied trips, compliance violations, or trust erosion.

I led the design standardization of how riders disclose service animals across multiple countries.

Instead of building country-specific exceptions, I defined:

• A consistent disclosure model

• Clear rider-facing language patterns

• Driver context surfaces at the right moment

• Localization rules that preserved structure without breaking compliance

This shifted the experience from fragmented implementations to a scalable foundation.

What changed

Before:

• Market-specific implementations

• Inconsistent rider expectations

• Driver confusion

After:

• Unified disclosure states

• Predictable interaction patterns

• Clear driver preparation context

• A model reusable in new countries

Impact (redacted)

• ↓ Service denial rates (double-digit % reduction)

• ↑ Trip completion reliability

• ↓ Accessibility-related support contacts

• Scaled to multiple new markets without redesign

Module 2: Hearing and Vision Self-Identification

Blind, low-vision, deaf, and hard-of-hearing riders often relied on manual messaging or calls to coordinate pickup.

This introduced friction at the most failure-prone moment of the trip.

I introduced proactive self-identification within the product, shifting communication from ad-hoc to structured.

The solution:

• Allowed riders to signal needs before arrival

• Integrated contextual guidance for drivers

• Reduced reliance on manual coordination

• Maintained rider control over disclosure

This was not just a new toggle — it was a systemic redesign of how pickup context flows between rider and driver.

System Value

• Structured communication replaces improvisation

• Driver preparedness improves reliability

• Reduced uncertainty during arrival

• Designed for dignity, not exception handling

Impact

• ↑ Pickup success rate for Blind/Low-vision riders

• ↓ Arrival confusion incidents

• ↑ Improved qualitative trust feedback

Module 3: Operationalizing Pickup Support

Accessibility-related pickup failures often came from unstructured rider requests.

I led multiple initiatives that transformed ambiguous, free-form needs into structured, in-context driver guidance.

This included:

• Translating rider preferences into actionable signals

• Structuring pickup and dropoff feedback loops

• Improving marketplace health while supporting mobility constraints

Because these systems touch internal marketplace logic and sensitive measurement frameworks, detailed flows are protected.

What this unlocked

• Shift from proxy metrics to experience-level measurement

• Reduced friction at high-risk pickup moments

• Improved operational clarity for drivers

・・・・・・・・・・・

System Impact

This body of work established accessibility as infrastructure rather than exception handling.

Across initiatives, we achieved:

• Measurable reductions in accessibility-related trip failures

• Improved rider trust signals

• Reduced operational ambiguity

• Increased scalability across markets

• A reusable design language for future accessibility efforts

All sensitive quantitative metrics have been intentionally redacted from this public version.

・・・・・・・・・・・

What This Enabled

The long-term value of this work wasn’t only feature improvement.

It enabled:

• Faster launches in new countries

• Consistent regulatory compliance

• Reduced design rework across teams

• A shared vocabulary for accessibility decisions

• A foundation extendable to emerging technologies (including AV experiences)

Accessibility became a system — not a patch.

・・・・・・・・・・・

Key Learnings

1. Accessibility at scale requires governance, not just empathy.

2. Structured context reduces friction more effectively than reactive messaging.

3. Marketplace dynamics must be designed into the system from the start.

4. Measurement models define what teams are capable of improving.

5. Systems thinking creates leverage far beyond a single feature.

・・・・・・・・・・・

🔒 Additional Details

Certain modules within this program involve internal marketplace logic, regulatory nuances, and sensitive performance metrics.

A deeper walkthrough — including detailed flows, measurement dashboards, and country-specific adaptations — is available upon request.