AISA - Proactive AI Defense
Combating the RM14.3 Billion Scam Crisis through Predictive Design
Recognition
MyHackathon Finalist (Top 4 Nationwide). Supported by Cradle Fund & Malaysia Govt.
Role
Lead Product Designer (UX/UI & Pitch Strategy) & AI Logic Orchestrator.
📌 The Strategic Vision
The Innovator
Most security tools are reactive—they tell you that you’ve been scammed after the money is gone. I identified that the "Vulnerability Gap" in Malaysia lies in the first 30 seconds of a scam call, where panic overrides logic.
The Mission
To move from recovery to prevention by creating an AI "Guardian" that intervenes in real-time.
🏛️ Crisis Mapping & Problem Logic
The Strategy
I approached this not just as an app, but as a defensive shield for the most vulnerable citizens.
- The "Reactive" Failure: Current solutions rely on manual reporting. By the time a number is blacklisted, the damage is done.
- The OSINT Integration: I architected a flow where the app uses Open Source Intelligence (OSINT) to verify incoming calls and messages against global threat databases instantly.
- Multi-Language Accessibility: Recognizing Malaysia’s diversity, I designed the UI to support voice and text in multiple languages, ensuring that a "Mak Cik" in a rural area receives warnings she actually understands.
🎨 Experience Design & Real-Time Intervention
The Designer Eye
Design was used as a psychological intervention to break the "Scammer’s Spell."
The "Panic-Break" UI
When a suspicious call is detected, the app triggers a high-contrast visual alert that provides a "Pause" moment, giving the user a clear "Safe" or "Danger" verdict.
Automated Evidence Analysis
To solve the friction of reporting, I designed an Evidence Upload Funnel. Users can take a screenshot of a suspicious chat, and the AI automatically parses the text to generate a formal report.
The Community Dashboard
Designed a real-time heat map of scams. If a neighborhood is being targeted, AISA sends a community broadcast to warn neighbors.
⚙️ Technical Governance & Scalability
The Architecture
Clean Architecture
Allows the system to swap AI models or add new security layers without a total rebuild.
12-Month Pilot Roadmap
Strategic rollout starting with senior citizens, allowing for real-time feedback loops to refine predictive accuracy.