PongoFin Documentation

Platform architecture

PongoFin transforms raw banking data into structured financial intelligence through a layered processing pipeline. Every transaction is validated, normalized, and classified before entering any calculation or report.

Processing pipeline

1. Import — Acquisition of bank statements and financial records from CSV, API, or manual entry
2. Normalization — Standardization of formats, dates, and descriptions
3. Classification — Assignment of each transaction to an accounting category
4. Reconciliation — Cross-referencing bank movements with accounting records
5. Analysis & Reporting — Generation of financial indicators, forecasts, and reports

Core modules

Supported data sources

transaction {
    date
    amount
    description
    counterpart
    category
}

Calculation methods

All financial calculations follow standard accounting principles, with additional analytical layers for projections and insights. Rules are deterministic and every change, classification, and reconciliation is tracked with a timestamp and user attribution.

Data normalization

During import, every record goes through:

• File parsing — Structure detection and field extraction
• Date conversion — Unified ISO format across all sources
• Currency handling — Conversion and normalization to base currency
• Description cleanup — Noise removal, abbreviation expansion
• Operation type identification — Credit / debit classification

Bank reconciliation

Reconciliation verifies consistency between bank movements and accounting records. The reconciliation engine applies multi-attribute matching:

Unmatched items are flagged for manual review, ensuring complete auditability.

Chart of accounts

Every transaction is assigned to an accounting category based on a configurable chart of accounts. Standard categories include:

• Revenue — Sales, service fees, interest income
• Operating costs — Rent, salaries, supplies, utilities
• Taxes — VAT, income tax, social contributions
• Investments — Capex, asset acquisitions
• Internal transfers — Inter-account movements

P&L statement

The profit and loss statement is calculated by aggregating classified transactions over a period:

  Revenue
− Operating costs
− Financial charges
− Taxes
─────────────────
= Net profit

Balance sheet

Financial forecasting

The platform generates forward-looking projections based on:

• Historical cash flow analysis — Past revenue and cost patterns
• Time trends — Growth rates, decay curves, momentum
• Projection models — Statistical techniques applied to historical data

Automated business plan

Data integrity

• Data validation — Schema and format checks on every imported record
• Balance verification — Ensures debits and credits remain balanced
• Duplicate detection — Identifies and flags potential duplicate transactions
• Cross-report consistency — Validates totals match across different report views
• Audit trail — Every change recorded with timestamp and user attribution

Categorization system

The system uses a 3-level hybrid approach to categorize transactions: deterministic rules, an AI-assisted model, and user review. Categories follow a hierarchical schema with parent categories and sub-categories.

The 3 classification levels

Hierarchical category structure

Matching rules

Deterministic rules evaluate:

• Counterpart — Entity/company name in the payment reference
• Description pattern — Keywords and regex on the bank transaction note
• Amount range — Typical amounts by category (e.g. recurring payroll)
• Frequency and periodicity — Recurring transactions at a fixed amount
• SWIFT/BIC code — Counterpart bank identification

Input and output examples

This section shows concrete examples of how the platform processes bank data, from the import phase through to final reports.

Bank statement CSV format

The expected input format for bank statements:

Date;Description;Amount;Currency;Balance
2025-01-03;SALARY CREDIT;+3200.00;EUR;5400.00
2025-01-05;RENT PAYMENT;-1200.00;EUR;4200.00
2025-01-07;ENEL UTILITY BILL;-180.50;EUR;4019.50
2025-01-10;INVOICE RECEIPT 2025-001;+4500.00;EUR;8519.50
2025-01-15;VAT F24 PAYMENT;-890.00;EUR;7629.50
2025-01-20;TRANSFER TO SUPPLIER XYZ;-560.00;EUR;7069.50
2025-01-25;EXPENSE REIMBURSEMENT;+120.00;EUR;7189.50

Output: classified transactions

Output: monthly P&L

Bank reconciliation — example

Export and output formats

PongoSID — Financial Accounts Registry

PongoSID is the module dedicated to preparing and submitting reports to the Financial Accounts Registry (Anagrafe dei Rapporti Finanziari) for obligated entities, professional firms, and financial intermediaries.

Who it's for

• Banks and financial intermediaries
• Fiduciaries and trust companies
• Professional firms (accountants, labor consultants)
• Obligated parties under D.P.R. 605/1973 and related regulations
• Operators managing ARF reporting on behalf of clients

Core features

Operational workflow

1. Entry or import of financial account data
2. Automatic validation of mandatory fields and formats
3. Pre-submission review with anomaly flagging
4. SID flow generation in the correct format
5. Archive of the report with receipt and date

Contact and support

For technical questions, access requests, or platform support, write to: