Bridging the gap between deep-tech AI signals and social media phishing and scam threats.

PhishSpot.ai is an AI-powered trust layer that detects fake profiles, impersonation accounts, and social media phishing threats — built for everyday creators, influencers, and small brands on Instagram, TikTok, and Facebook.

I owned the end-to-end product design and foundational brand direction.

My Role(s)

Brand & UI/UX Designer

Tools

Figma, Illustrator, Photoshop

Industry

AI Security/B2C

View Site

The Problem

Identity theft has moved beyond email. It's thriving on social media and most victims only find out after the damage is done. But existing security tools made things worse, not better.

Too clinical

raw data dumps with no context, leaving users overwhelmed instead of empowered.

No clear next step

tools could detect a threat but couldn't tell users what to do about it. "Okay, I'm being impersonated... now what?"

Built for the wrong audience

enterprise-grade security language and interfaces, handed to everyday creators and small business owners with no security background.

My Role
And Approach

I designed the full product experience across:

  • Profile scanning and lightweight onboarding flows
  • Impersonation detection and identity verification insights
  • Threat analysis dashboards for tracking fake accounts and suspicious activity
  • Scan results interpretation and guided next-step recommendations
  • Core conversion flows for free scans and early access acquisition
  • Full brand identity; logo, visual language, and typography

My goal was to simplify complex AI-driven security signals into a clear experience that helps users answer three questions: Who is real? Who is fake? What do I do next?

The Process

Zero-barrier entry: Security onboarding is notoriously tedious. I stripped it down to the minimum users drop in a social handle or reference image and scanning begins immediately. No lengthy setup, no account required to start.

The waiting state as a trust-builder: AI processing takes time, and blank loading screens kill confidence. I designed an interactive in-progress state that visually maps the backend analysis in real time, metadata checks, behavioral patterns, image mismatch detection, turning a wait into a credibility-building moment that shows users the system is actually working.

Human-readable threat results, not raw data: Once the scan completes, the dashboard translates technical signals into a clear, human-readable risk summary. No jargon. No raw telemetry. Just a straightforward answer to: is this account a threat?

Action over diagnosis: The UI doesn't stop at identifying threats. It drives immediate resolution, auto-generating platform takedown reports, alerting followers, and enabling continuous automated monitoring with a single action.

The
Results

My Output

  • 25+ screens across web and responsive mobile
  • 4 core user flows - scanning, detection, dashboard, and conversion
  • Full brand identity - logo, typography, colour system, and design system
  • 3 platforms supported in the scan flow - Instagram, TikTok, and Facebook

Impact

A perfect Cumulative Layout Shift score of 0 means every element lands exactly where it should, nothing shifts or jumps as the page loads. That's a direct result of intentional layout and spacing decisions made at the design stage.

A 93 accessibility score reflects design choices made for real inclusivity, colour contrast, typography sizing, and visual hierarchy that works for everyone, not just ideal users.

With PhishSpot.ai, we successfully transformed complex, invisible AI security signals into an intuitive defensive shield for everyday creators and brands. By prioritizing progressive disclosure and clear, actionable mitigation steps over technical anxiety, the design successfully shifted digital identity protection from a reactive panic into proactive, effortless control.

You can also see...

Apperr Website, 2024-2025

View Project

TrialClinIQ, 2025

View Project