Skip to content

CSA Loom — the Microsoft Fabric experience for Azure tenants where Fabric isn't yet available: lakehouses, warehouses, notebooks, semantic models, Activator rules, Data Agents, across Commercial, GCC, GCC-High, and DoD IL5

Synthetic financial transactions dataset (CUI-safe)¶

A fabricated retail-banking transaction set for FSI / regional-bank workshops. Used as the Day-3 alternate workload and for fraud-pattern KQL exploration.

CUI-safe by construction

All account numbers, customer IDs, merchants, and amounts are machine-generated. No real financial records, account numbers, or PII are present. Account IDs are obviously synthetic (ACCT-…) and never resemble real card/account formats.

Files¶

File	Purpose	Rows (workshop size)
`transactions.csv`	Transaction fact (Bronze source)	~80,000
`accounts.csv`	Account dimension	~2,000
`merchants.csv`	Merchant dimension	~300

Schema — `transactions`¶

Column	Type	Notes
`txn_id`	string (UUID)	Synthetic surrogate key
`account_id`	string	FK → `accounts.account_id` (e.g., `ACCT-018273`)
`txn_timestamp`	timestamp (UTC)	Spread across ~90 synthetic days
`merchant_id`	string	FK → `merchants.merchant_id`
`amount`	decimal(12,2)	Plausible spend; ~1% injected fraud-like outliers
`channel`	string	`card_present` \| `online` \| `transfer` \| `atm`
`country`	string	Generic 2-letter code; some mismatched for fraud labs
`is_flagged`	boolean	Synthetic fraud label (ground truth for the lab)

Schema — `accounts`¶

Column	Type	Notes
`account_id`	string	PK (synthetic)
`account_type`	string	`checking` \| `savings` \| `credit`
`open_date`	date	Synthetic
`home_country`	string	Generic code (drives the country-mismatch signal)

Schema — `merchants`¶

Column	Type	Notes
`merchant_id`	string	PK
`merchant_category`	string	MCC-like label (e.g., `grocery`, `travel`)
`risk_tier`	string	`low` \| `medium` \| `high` (synthetic)

Sample rows — `transactions`¶

txn_id,account_id,txn_timestamp,merchant_id,amount,channel,country,is_flagged
e5a1...,ACCT-018273,2026-05-02T13:22:00Z,MER-0142,42.10,card_present,US,false
f6b2...,ACCT-018273,2026-05-02T13:31:00Z,MER-0307,2980.00,online,RO,true
07c3...,ACCT-004511,2026-05-03T08:05:00Z,MER-0211,11.75,atm,US,false

Lab use¶

Day 3 (Transform, FSI alternate): Bronze → Silver normalizes channels + country codes; Silver → Gold builds gold.account_daily_spend and a fraud-candidate view keyed on is_flagged + country mismatch.
Day 3 (KQL): ingest transactions into ADX and run a fraud-pattern query (velocity + country mismatch) producing a candidate list.
Day 4 (Data Agent): ground an agent on the Gold spend tables for natural-language questions ("largest flagged transaction last week").