Zalingo Synthetic Finance — Fraud & Chargebacks (CNP) — 100k

Need more volume or notebooks? Upgrade to the Premium Evaluation Kit (~1M rows) or ask for a weekly/daily enterprise feed.

Dataset Features (representative)

• Core: transaction_id, account_id, ts_utc, amount, currency, channel (ecommerce | wallet | mail/phone), merchant_id, mcc, merchant_country, user_agent
• Auth Signals: three_ds_result, avs_result, cvv_result, auth_result, decline_reason_code
• Device/IP/Geo: device_fingerprint, ip_country, distance_km_billing_shipping, first_time_merchant_flag, recurring_flag, coupon_used
• Velocity (enriched): txn_ct_15m/1h/24h/7d, amount_sum_1h/24h, unique_merchant_ct_7d
• Labels & Scores: fraud_label (0/1), chargeback_flag (0/1), risk_score_0_1 (Columns may vary slightly; see the included preview for exact schema.)

Distribution

• Format: ZIP containing Parquet data (100k_sample.parquet), sample_100.csv (preview), and schema.json
• Volume: 100,000 rows, ~22–35 columns
• Approx Size: 3–6 MB zipped (mix-dependent)
• Structure: Single Parquet (or a few shards); schema stable across focused fraud packs

Usage

• Fraud/Risk modelling: baseline models, feature ablations, threshold & policy tuning
• Authorization optimisation: AVS/3DS strategy experiments and trade-off analysis
• Scenario testing: velocity, geo-mismatch, first-use & recurring patterns
• MLOps QA: schema contracts, drift monitors, dashboard demos

Coverage

• Geographic: Multi-country synthetic coverage (ISO codes)
• Time Range: Recent multi-year synthetic window with weekly/seasonal patterns
• PII: None — fully synthetic; not re-identifiable

Who Can Use It

• Risk/Data Science — feature engineering & model iteration
• Payments/FinOps — authorization strategy & loss-rate diagnostics
• Product/Analytics — KPI sandboxes & experiment design
• Vendors/SIs — demo environments & connector validation

Notes / Disclaimers

• Not real cardholder data. Not for production credit decisions.
• Rates, labels, and distributions are synthetic and calibrated; they do not represent any specific issuer/acquirer/PSP.