Cisco Directory · 2025
AI-powered profile picture validator
Improve the Cisco Directory photo upload experience by using AI to validate guideline compliance in real time, reducing approval delays, manual review effort, and user frustration.
My role
Lead product designer
Project type
Product design & strategy
Year
2025
Company
Cisco
Problem & solution
Clear feedback should come before rejection
Employees often submit profile photos that do not meet Cisco’s guidelines, resulting in manual rejections, delayed approvals, and repeated resubmissions. Feedback is provided only after review, creating frustration for users and significant workload for the Corporate Security & Badging team. The process is inefficient, slow, and does not scale.
The users

Security & Badging
Reviewing and approving submissions.

Employees
Uploading profile photos.
Constraints
Two weeks. A small team. Focused improvements.
With only two weeks and a small team, myself and a content designer, the project focused on targeted improvements rather than a full redesign. By leveraging existing Directory patterns, we delivered meaningful usability enhancements quickly while maintaining a familiar and seamless experience for users.
Design highlights

A square crop for a circular profile photo
Before
The crop tool used a rectangular frame that did not match the circular profile photo display, creating confusion about how the final image would appear.
After
The crop experience was updated to a square frame, aligning with common profile photo standards and making the final result clearer and more predictable.
Research & inspiration
Learning from other photo-validation experiences
To kick off the project, I explored how other companies incorporated AI into validating user-uploaded photos to identify common patterns and best practices. I found two relevant references: Airbnb and 24 Hour Fitness.
Success metrics & reflection
How we’ll measure success
We’ll track how this design performs and decide on next steps using the established success metrics. These are targets, not reported outcomes.
Reduce manual rejections by 50% within six months.
Achieve a first-time approval rate of over 90%.
Improve user satisfaction to an average score of 4 out of 5 or higher.
Complete the AI validation process within an average of 3 seconds per image.
Small changes can make a meaningful difference
This project reminded me how much impact small, focused changes can have. Working with tight timelines and limited resources taught me to prioritize what matters most for users, like clear feedback and intuitive interactions, while still making AI feel helpful and easy to use.
←
Explore more work