SCAMRADAR+
Trusted by 60K+ users protected.
Capabilities & Performance
WHAT IT DOES & HOW WELL
Six layers of AI-powered protection, validated on 46,360 real-world messages.
AI Scam Detection
A Calibrated Logistic Regression model trained on 46,360 real messages scores every input and returns an instant scam or ham verdict.
Phishing URL Analysis
Any URLs embedded in a message are extracted and checked against live reputation feeds to catch phishing links before they do damage.
Vector Pattern Matching
FAISS nearest-neighbour search surfaces the closest known scam patterns from the training corpus so you can see exactly what it resembles.
Real-Time Response
The full verdict — score, label, and similar matches — is returned in under 200 ms via a single FastAPI endpoint.
Multi-Channel Coverage
Works on SMS, email, WhatsApp, or any plain-text input. No channel-specific retraining required.
Smart Risk Scoring
A 0–100 confidence score explains exactly how certain the model is, letting you set your own risk threshold.
0.00
%Model Accuracy
+16.39% from v1 baseline
0.00
%Precision Score
+2.47% from v1 baseline
0.00
%Recall Score
+2.12% from v1 baseline
0.00
%F1 Score
+13.3% from v1 baseline
Confidence Separation
Score deviation from decision threshold (t = 0.47)
Precision–Recall Balance
P − R gap across threshold sweep (zero = balanced)
Training Convergence
Accuracy gain Δ% per gradient step
Latest Scam Trends
“Impersonates major banks with urgent account suspension alerts. Links redirect to credential-harvesting portals hosted on lookalike domains registered hours before the campaign.”
Fake Bank SMS Scam
🔴 Critical · SMS Phishing
“Spoofed PayPal emails claiming account limitation. Lookalike domains collect login credentials and 2FA codes in real time and relay them to attacker-controlled servers.”
PayPal Phishing Campaign
🟠 High · Email Phishing
“Fake celebrity-endorsed crypto giveaways spreading via WhatsApp and Telegram, requesting a small 'verification' transfer to unlock fabricated winnings.”
Crypto Giveaway Scam
🟠 High · Social Engineering
“Carrier impersonation SMS (DHL, FedEx) with 'package held' alerts. Tracking links lead to malware-laced pages that harvest device credentials and banking details.”
Fake Delivery Tracking
🟡 Medium · SMS Phishing
“Fake job offers via LinkedIn and email promising remote positions. Victims are asked to pay an upfront 'equipment fee' or share banking details for payroll setup.”
Fake Job Offer Scam
🟠 High · Advance Fee Fraud
Trusted by security-conscious users
Real feedback from people who use ScamRadar+ to protect themselves and their teams from scams every day.
"We integrated ScamRadar+ into our customer support pipeline. It flags suspicious messages before they reach agents and has already blocked several social engineering attempts. The API response time is impressively fast."
Lior Ben-David
IT Security Lead, FinTech IL
"The FAISS similarity search is a clever addition — seeing which known scam a message resembles gives analysts real context, not just a binary label. The 97% accuracy on the test set holds up in practice."
Noa Shapiro
Cybersecurity Researcher, Tel Aviv University
"We evaluated three scam-detection APIs before choosing ScamRadar+. The combination of ML scoring, vector pattern matching, and live URL scanning in a single call made it the obvious choice for our product."
Daniel Katz
Product Manager, SafeComm
Trusted by people who can't afford to be wrong
ScamRadar+ · 2026
MEET THE TEAM
Behind ScamRadar+
Team
Lead Developer & ML Engineer
Ameer Hassouna
"Built the full ML pipeline, FastAPI backend, and Next.js frontend. Passionate about making AI tools that actually protect people."
FAQ
Frequently Asked Questions
Everything you need to know about ScamRadar+. If you don't find the answer you're looking for, feel free to reach out.
Can't find what you're looking for? Contact our support team