Audit & Technology
E-Ledger Analysis Engine
An analysis engine for examining Turkish e-ledger (e-Defter) data in terms of account movements, unusual transactions and risk indicators.
Built by: Ufuk Süleyman Oflaz
The problem
The e-ledger holds every accounting movement of a company in structured form, yet the data is usually kept only to meet a legal requirement. Finding what is unusual among tens of thousands of journal lines is slow and incomplete with traditional, sample-based review.
Why it matters
An unusual transaction often looks insignificant on its own; its meaning only emerges when it is seen within the full population. Sample-based review is precisely the method most likely to miss such transactions.
Approach
- Read e-ledger files and turn them into a data model that can be analysed at account, voucher and line level.
- Rebuild the trial balance from movements and track account balances through the period.
- Flag unusual transactions with defined indicators and rank them by review priority.
- Move the review from a sample to the full population.
What it examines
- Account movements and debit/credit analysis
- Trial balance generation and balance tracking
- Detection of accounts with reverse balances
- Flagging of repeated and round amounts
- Entries posted on weekends and holidays
- High-value and unusual entries
- Relationships between journal entries
- Summary of risk indicators
Example output
Counts of entries flagged by each indicator, monthly entry volume, accounts with reverse balances and journal lines ranked by review priority.
Limitations
- A flagged entry is not necessarily wrong or improper; it only indicates that it should be reviewed.
- Indicator thresholds must be calibrated to the company’s business.
- The analysis depends on complete and accurate e-ledger data.