UX Design · Amazon Web Services · 3 years
AWS Project Libra —
Hybrid Meeting System
Designed the inclusive, intelligent meeting room system that bridges in-room and remote participants. 477+ meetings deployed globally across AWS offices, 13% SCAT improvement representing ~$2.8M in influenced sales pipeline.
The Problem
Remote participants could see and hear fine. They couldn't participate fine.
After shadowing 15 team meetings in the first month, I found a clear pattern: in a 10-person meeting with 7 in-room and 3 remote, in-room participants made eye contact, read body language, and jumped into conversations fluidly. Remote participants watched a static wide-angle shot of the room, waited for awkward pauses, and were frequently talked over. The meeting effectively ran as if the remote people weren't there.
Tension 1
Equity vs. Adoption
Individual cameras solved equity but created in-room discomfort. If we removed cameras, we'd lose the equity we were trying to create.
Tension 2
Remote vs. In-Room NPS
Pilot showed remote participants loved individual tiles (+40 NPS) while in-room participants felt "watched" and distracted (-20 NPS).
Tension 3
AI vs. Privacy
83.9% wanted AI summaries, but 13 users explicitly rejected engagement analysis as "frightening" and "invasive."
Deep Dive
Designing the AI meeting agent
With only two months to deliver a working demo, the challenge was not just introducing AI into meetings — but determining when it should speak, and when it should stay silent. This was 2023. Every competitor was shipping AI engagement features. Leadership asked why we weren't leveraging ML capabilities.
Key constraints I navigated
- Designing within an existing component library to ensure speed and consistency
- Integrating summaries from third-party platforms (Zoom, Teams, Chime)
- Connecting to user identity and authorization models
- Filtering non-essential information to prevent cognitive overload
- Designing with neurodiverse participants (ADHD, autism) in mind from day one
Decision approach
Given the compressed timeline, I focused on a grounded experience rather than speculative AI behaviors, emphasizing clarity, relevance, and participant control. This meant intentionally rejecting several ideas that introduced unnecessary automation or surfaced information without strong context.
"This is better than our OP1 executive rooms. Can we deploy in all leadership spaces?"
— VP endorsement post-demo
Impact
Business outcomes
477+Meetings deployed globally across AWS offices
3,151+In-room participants over 3.5 years
13%SCAT score improvement (~$2.8M influenced pipeline)
85+Beta program NPS maintained
Three critical product decisions
- Removed self-view by default — 60% reduction in "feeling watched" complaints
- Killed 16-tile grid despite customer requests — research showed it caused split attention and hurt neurodiverse users
- Built the AI meeting agent — demoed and shipped in 2 months, NPS stayed above 85
Accessibility impact
- Led accessibility audit — Born Accessible compliance from day one, not afterthought
- Designed for neurodiverse users (ADHD, autism), low vision, deaf/hard of hearing, and physical disabilities
- 100% of accessibility users preferred Libra over traditional conference rooms
- Customizable captions, content zoom/pinch, consistent information placement became core product features
Design system
Scaled the enterprise design system to 400+ responsive components across Web, Mobile, and TV surfaces. Reduced engineering implementation time and ensured WCAG compliance across the entire product suite.