PROBLEM STATEMENT ❓

During year-end audits, a significant volume of ledger accounts and voucher entries must be scrutinized—even after sampling techniques are applied. This process is further complicated by limited time, large datasets, and increasingly complex regulatory frameworks.

SOLUTION ✨

STEP 1: 🔠

Define the Input:

In this use case, the primary inputs are the Trial Balance and the Columnar Day Book

STEP 2 & 3: 🧹

1) Standardize and Classify the Input:

i) Data Standardization: Clean the input data to conform to a consistent format.

(e.g., removing subgroup headings interspersed among ledger entries)

ii) Data Classification: Categorize the ledgers appropriately.

(e.g., grouping into Assets, Liabilities, Revenue, and Expenses Also, TDS sections & Subgroups). This enables the AI model to interpret and process the data more efficiently.

2) Process Logic Development:

i) Frame audit rules to automate scrutiny. These rules are capable of handling high-volume datasets and help in generating the required outputs.

ii) With the support of AI ✨, these rules (representing human logic) are transformed into executable machine code. This allows the process to run seamlessly on large datasets—delivering results almost instantly.

iii) Testing and Validation: Rigorously test the process across multiple data types—simple, complex, and error-prone datasets. Through iterative testing, we ensure the solution is accurate, reliable, and capable of flagging anomalies or inconsistencies in the input data.

STEP 4: ⚙️

1) Integrate and Execute the Process: Integrate the developed logic with Excel using the VBA editor.

i) Open Excel in which you have the cleaned & classified Trail Balance, press Alt+F11, and paste the VBA code press Ctrl +S and closing the VBA screen.

ii) Then press Alt+F8 in excel file to execute the code.

CONCLUSION:

The output is generated in the form of: A high-level summary sheet, and Detailed rule-based sheets (one sheet per criterion applied).

Note:

1) With the help of AI & Human supervision (Supervised Learning) we are frequently updating the VBA code as and when rules have to be updated to match the regulatory amendments. (Ex: Updation of TDS threshold of 194J(b) from 30,000(till FY 24-25) to 50,000(from FY 25-26))

2) In voucher scrutiny the TDS section mapping & Subgrouping mapping can be done using “VLOOKUP” formula in Input Daybook sheet by taking the trial balance as the array.

For Reference: LEDGER SCRUTINY SUMMARY SHEET