Start

Guardian Layer

A browser extension that protects older adults from scam emails, without asking them to learn anything new.

Project type
Browser extension, UI/UX design
Problem
Creating a product that protects elderly people from scams.
Impact
6 real users.
View live onboarding
“I’m always very careful when sending money to a new person.”
“I’ve gotten messages of a link to pay for package delivery, turns out it was a fake mailing website”
“My mom always asks me whether this is real and should I pay”
An older adult reading email on a laptop, the kind of user Guardian Layer is designed to protect.
The original Guardian Layer phone app: a home shield screen and a weekly summary of scams intercepted.
The phone app that couldn’t ship

Why it couldn’t be built

A phone app can’t read texts or monitor calls on iOS, the permissions ask is heavy, and the backend was bigger than one designer could ship.

So: a browser extension

Same protection, different form. A Gmail extension that flags scams where the threat lives. A smaller trust ask, and nothing new for the user to learn.

Session A

Tested how users react to the warnings.

Will a pre-view warning actually change behavior?

Yes, they read the warning first and decided to not act upon the email.

Session B

Tested how users react to the false warnings.

What happens to trust when the tool is wrong?

One type of group lost the trust while the other believed it was right.

The person it protects
The person who installs it
Needs warnings simple enough to trust without thinking.
Needs reasoning detailed enough to audit before they trust the system.
They have opposite needs for the same interface.

Design

One warning. Two readings.

View live onboarding

It names what's wrong, not just that something is wrong. Three specific reasons.

It tells the user what to do. Ends with a clear action.

The Guardian Layer warning component: three plain-language reasons an email looks suspicious, ending with what to do.

It speaks the user's language. No "phishing," "spoofed domain," or "malicious URL."

The warning component

It explains how Guardian Layer reads and evaluates emails.

The false positive policy is shown upfront, including how the system handles its own mistakes.

The full Guardian Layer onboarding flow, confirming protection for both users before showing the product.

It confirms that the protection has been set up.

It shows the product instead of describing it.

The privacy model is stated clearly: what gets read, what gets stored, and what stays untouched.

The onboarding design
Guardian Layer running in a real Gmail account, showing a scam warning above a suspicious email.

This started as a project about payment apps. It became a project about protecting people who don’t know they need it.

View full case study