FraudGuard AI – Credit Card Fraud Detection

Total

—

Approved

—

Flagged

—

Blocked

—

Avg Risk

—

Volume

—

\n\n\n\n\n

💳 Submit Transaction

\n Amount (USD)\n \n

\n Card Last 4 Digits\n \n

\n Merchant Name\n \n

\n Category\n \n

\n Time of Day\n \n

\n Location\n \n

\n 🌍 International Transaction\n \n

\n 🌐 Online / Card-Not-Present\n \n

\n \n

\n\n

📊 Live Insights

📈

Submit a transaction to see live insights

\n\n\n

\n Recent Transactions\n \n

\n \n \n \n \n \n \n

#	Merchant	Category	Amount	Card	Location	Time	Risk	Status	Date
📭 No transactions yet

\n\n\n

🧠 Detection Rules

💰

Transaction Amount

Transactions over $2,000 trigger medium risk (+20 pts); over $5,000 trigger high risk (+35 pts).

🌍

International Flag

Transactions made from international locations add +25 risk points.

🌙

Unusual Hours

Transactions between midnight and 5am are suspicious (+20 pts).

⚠️

High-Risk Categories

Cryptocurrency, gambling, and wire transfers add +25 pts each.

🌐

Online + International

Card-not-present combined with international adds +15 pts bonus.

📊 Risk Thresholds

✅ ApprovedScore: 0–39

Transaction proceeds normally with no intervention.

⚠️ FlaggedScore: 40–69

Transaction is held for manual review by a fraud analyst.

🚫 BlockedScore: 70–100

Transaction is automatically blocked and cardholder is notified.

\n 💡 Tip: Try submitting a transaction with amount $8,500 + Cryptocurrency + Midnight + International + Online to see a perfect 100/100 risk score blocked!\n

\n\n\n\n\n\n`);\n});\n\n// API: Submit transaction\napp.post(\"/api/transactions\", async (c) => {\n const body = await c.req.json();\n const { amount, merchant, category, location, cardLast4, timeOfDay, isInternational, isOnline } = body;\n\n const { riskScore, reasons, status } = detectFraud({\n amount, category, timeOfDay,\n isInternational: !!isInternational,\n isOnline: !!isOnline,\n });\n\n await sqlite.execute({\n sql: `INSERT INTO transactions (amount, merchant, category, location, card_last4, time_of_day, is_international, is_online, risk_score, status, flagged_reasons)\n VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)`,\n args: [\n amount, merchant, category, location, cardLast4, timeOfDay,\n isInternational ? 1 : 0, isOnline ? 1 : 0,\n riskScore, status, JSON.stringify(reasons),\n ],\n });\n\n return c.json({ riskScore, reasons, status });\n});\n\n// API: Get all transactions\napp.get(\"/api/transactions\", async (c) => {\n const result = await sqlite.execute(\n \"SELECT * FROM transactions ORDER BY created_at DESC LIMIT 100\"\n );\n return c.json(result.rows);\n});\n\n// API: Get stats\napp.get(\"/api/stats\", async (c) => {\n const total = await sqlite.execute(\"SELECT COUNT(*) as count FROM transactions\");\n const flagged = await sqlite.execute(\"SELECT COUNT(*) as count FROM transactions WHERE status = 'flagged'\");\n const blocked = await sqlite.execute(\"SELECT COUNT(*) as count FROM transactions WHERE status = 'blocked'\");\n const approved = await sqlite.execute(\"SELECT COUNT(*) as count FROM transactions WHERE status = 'approved'\");\n const avgRisk = await sqlite.execute(\"SELECT AVG(risk_score) as avg FROM transactions\");\n const totalAmount = await sqlite.execute(\"SELECT SUM(amount) as total FROM transactions\");\n\n return c.json({\n total: total.rows[0]?.count ?? 0,\n flagged: flagged.rows[0]?.count ?? 0,\n blocked: blocked.rows[0]?.count ?? 0,\n approved: approved.rows[0]?.count ?? 0,\n avgRiskScore: Math.round((avgRisk.rows[0]?.avg as number) ?? 0),\n totalAmount: Math.round((totalAmount.rows[0]?.total as number) ?? 0),\n });\n});\n\n// API: Clear all transactions\napp.delete(\"/api/transactions\", async (c) => {\n await sqlite.execute(\"DELETE FROM transactions\");\n return c.json({ success: true });\n});\n\napp.onError((err) => Promise.reject(err));\n\nexport default app.fetch;\n","deleted","label",[102],"permalink","https://www.val.town/x/Rubasri/credit-card-fraud-detection/code/main.ts","extension","actionData","errors"]