Why Bank Statement Analysis Matters
Bank statements are the primary source document for pre-filing financial analysis. They contain the complete transaction record — income receipts, expense payments, transfers, loan disbursements, and regulatory transactions. Accurate extraction and classification of this data is the foundation of reliable compliance preparation.
For accountants managing multiple clients across different banks, the challenge is structural: every bank uses different PDF layouts, column headers, date formats, and amount conventions. Manual data entry is slow and error-prone.
How Prefiler Analyses Bank Statements
Prefiler uses a profile-driven deterministic parsing approach. Each supported bank has a structured layout profile that defines:
- • Column header patterns — Keyword matching for date, narration, debit, credit, and balance columns
- • Date format recognition — Regex-based parsing for bank-specific date conventions
- • Polarity conventions — Whether amounts use separate debit/credit columns, Dr/Cr suffixes, or signed values
- • Footer pattern filtering — Automated removal of non-transaction content (totals, disclaimers, page numbers)
- • Numeric format handling — Indian grouping conventions (1,23,456.78)
Supported Banks
Tier 1
- • HDFC Bank
- • State Bank of India (SBI)
- • ICICI Bank
- • Axis Bank
- • Yes Bank
- • Kotak Mahindra Bank
- • Punjab National Bank (PNB)
Tier 2
- • IDFC First Bank
- • IndusInd Bank
- • Bank of Baroda
- • Canara Bank
- • Union Bank of India
From Extraction to Analysis
Once transactions are extracted, Prefiler automatically classifies each transaction using heuristic pattern detection. Categories include salary income, business receipts, GST payments, loan EMIs, investment transactions, and more.
The classified data feeds into the deterministic risk scoring engine, which evaluates transaction concentration, materiality thresholds, and cross-category patterns to produce a structured risk assessment.
Reconciliation Verification
Every extraction includes automated reconciliation: computed closing balance is verified against the statement's closing balance using integer arithmetic (eliminating floating-point errors). Any discrepancy is flagged with the exact variance amount.
Bank Statement Analysis Pipeline
- Upload PDF bank statement
- Automatic bank identification via layout matching
- Profile-driven deterministic extraction
- Transaction classification with confidence scoring
- Reconciliation verification against closing balance
- Risk scoring and structured report generation
Learn more: document extraction technology, pre-filing analysis workflow, technical architecture.