Automating Balance Confirmation request Letters using AI
Author: CA. Anshu Agarwal
Author: CA. Anshu Agarwal
Auditing is a critical function in the field of Chartered Accountancy. One essential part of the audit process is obtaining balance confirmation letters from external parties, such as debtors and creditors of our auditee client. Traditionally, this task involves preparing and sending individual letters to each party, which is both time-consuming and susceptible to human error, especially when the number of parties is large.
In this used case, I demonstrate how I used AI-powered tools and automation to completely streamline this task, turning it into an efficient, error-free process. The use case is based on real-life implementation in my practice and was developed with guidance from ChatGPT.
To automate the generation, personalization, and delivery of request letter for balance confirmation during audits using commonly available tools such as MS Word, Excel, and Python—with the help of AI guidance from ChatGPT.
Tools & Technologies Used
- Microsoft Excel (for data storage and tracking)
- Microsoft Word Mail Merge (for letter generation)
- Python (for PDF processing and email automation)
- ChatGPT (for step-by-step assistance and script generation)
Workflow Overview
The automation process was divided into three main steps:
- Step 1: Generating letters using Mail Merge & generate a single pdf file for all letters
- Step 2: Splitting the merged PDF into individual letters
- Step 3: Sending the letters via automated emails
The first step was to use Microsoft Word’s Mail Merge feature to automatically generate balance confirmation letters. With the guidance of ChatGPT, I created an Excel file containing the required details such as name, address, and email of each recipient.
Using Word’s Mail Merge, I connected this Excel sheet to a pre-designed letter template. The fields in the template were automatically populated for each recipient, and all the letters were generated into a single PDF file. This eliminated the need to manually create each letter.
Step 2: Splitting the PDF into Individual Files using Python
Once the consolidated PDF file was generated, the next challenge was to split this file into individual PDFs—one per recipient—and rename them properly.
Again, with assistance from ChatGPT, I used a Python script that performed the following:
- Read the recipient names from the Excel file
- Split the PDF into separate pages
- Rename each output file based on recipient data
This automation not only saved hours of manual effort but also ensured accuracy in naming and organizing the documents.
This step can also be used for separation of other documents also, like bulk bill generated from tally or other documents etc.
Step 3: Automating Email Delivery with Python
The final step was to send out the personalized balance confirmation letters to each recipient. To avoid manual emailing, I automated this task using Python.
With a script provided by ChatGPT, I accomplished the following:
- Read email addresses from Excel
- Attach the corresponding PDF file to each email
- Customize the email body
- Send the emails using SMTP
The result was a fully automated email process that saved significant time and reduced the chances of mistakes.
Implementing this AI-driven solution has brought multiple benefits to my practice:
- Eliminated repetitive and manual tasks
- Reduced human errors significantly
- Improved turnaround time for audit processes
- Increased productivity and allowed staff to focus on high-value tasks
Beyond balance confirmations, this workflow can also be adapted for:
- Splitting and mailing invoices from Tally
- Sending tax filing reminders
- Circulating client communications in bulk
This project showcases the power of combining domain expertise with AI assistance. By using accessible tools and AI-guided automation, I have demonstrated how Chartered Accountants can modernize traditional processes. The success of this use case opens doors for many more applications of AI in the practice of accountancy.