Skip to content

Treasury Data - Current Solutions & Workarounds

Tools & Systems

Airflow (Ingestion Pipelines)

  • What it is: Workflow orchestration for scheduled and event-driven data ingestion
  • How Palm uses it: DAGs for ingesting bank statements from S3/GCS/Azure Blob, SFTP, HTTPS APIs
  • Key capability: Freshness sensors detect when expected data hasn't arrived

Python Parsers (Format Parsing)

  • What it is: Custom parsers for standard and non-standard bank statement formats
  • Formats supported: ISO 20022 CAMT, MT940/MT942, BAI2, CSV with custom extensions
  • Key capability: Multi-format harmonization into unified schema

PostgreSQL (Data Storage)

  • What it is: Core data store with palm schema for normalized transaction and balance data
  • Key capability: Entity resolution, multi-currency handling, value-date logic

DBT (Data Transformation)

  • What it is: Silver/gold layer transformation models for analytics-ready data
  • Key capability: Schema validation, data quality checks, audit trails

How Teams Handle Data Today (Without Palm)

ERP as Source of Truth

  • What they do: Treat ERP (SAP, Oracle, PeopleSoft) as authoritative source for balances
  • Limitation: Only accurate after reconciliation (often monthly)
  • Used by: On, IHG

Manual Reconciliation

  • What they do: Monthly process to match bank statements against ERP records
  • Limitation: Creates a "trust gap" — data between reconciliations is approximate
  • Used by: On (outsourced to Deloitte)

Looker/BI Dashboards

  • What they do: Aggregate ERP data into visual dashboards for cash by currency
  • Limitation: Only as fresh as the underlying ERP data
  • Used by: On