Step 2 of 5
Field Mapping
Map the columns detected in your CSV to the canonical Capital AI transaction schema. Required fields must be mapped before training can begin.
Mapping completeness
82%
14 of 17 canonical fields mapped
Required fields
6/6
All required fields covered
Detected columns
23
14 mapped · 9 unused
Detected columns
From the uploaded CSV
- Mapped
txn_id
string · sample: txn_8a3f10c
- Mapped
event_ts
datetime · sample: 2024-04-12T18:42:11Z
- Mapped
cust_ref
string · sample: cus_44211
- Mapped
card_token
string · sample: card_4f9b
- Mapped
ip
string · sample: 84.122.18.4
- Mapped
merchant
string · sample: mer_12044
- Mapped
mcc
string · sample: 5411
- Mapped
country
string · sample: US
- Mapped
city
string · sample: Brooklyn
- Mapped
ch_type
string · sample: CNP
- Mapped
amt
number · sample: 84.21
- Mapped
ccy
string · sample: USD
- Mapped
label_fraud
boolean · sample: false
- Mapped
fraud_kind
string · sample: —
- Unused
auth_code
string · sample: A2918
- Unused
cvv_result
string · sample: M
- Unused
avs_result
string · sample: Y
- Unused
three_ds
string · sample: challenged
- Unused
issuer_resp
string · sample: 00
- Unused
device_fp
string · sample: —
- Unused
lat
number · sample: —
- Unused
lng
number · sample: —
- Unused
session_id
string · sample: ses_99af1
Canonical transaction schema
Select a source column for each field
transaction_id
Requiredtimestamp
Requiredcustomer_id
Requiredcard_id
Requireddevice_id
ip_address
merchant_id
Requiredmerchant_category
merchant_country
merchant_city
merchant_latitude
merchant_longitude
transaction_type
amount
Requiredcurrency
is_fraud
fraud_type
