1. Executive Summary

GST reconciliation is one of the most time-consuming and repetitive activities in accounting and taxation practice. Chartered Accountants and audit professionals frequently work with large volumes of transactional data received from Books of Accounts, GSTR-2A, GSTR-2B, Sales Registers and other GST-related reports.

GST Insight is a Python-based GST reconciliation and audit intelligence platform developed to simplify, automate and optimize reconciliation workflows. The application combines structured data processing, database-driven matching logic and an interactive review dashboard to assist professionals in identifying mismatches, missing invoices, duplicate entries and tax variations efficiently.

The platform uses Pandas for data preprocessing, SQLite as the reconciliation engine and Streamlit for the interactive review interface. The primary objective of the platform is not to replace professional judgement, but to assist professionals by reducing repetitive reconciliation work and improving review efficiency.

2. Problem Statement

Traditional GST reconciliation methods mainly depend on spreadsheets, filters and manual verification. As the transaction volume increases, the reconciliation process becomes difficult, time-consuming and error-prone.

Professionals commonly face practical challenges such as:

• Duplicate invoice entries

• GSTIN mismatches

• Date inconsistencies

• Invoice number variations

• Missing invoices between Books and GST reports

• Large Excel file handling issues

• Difficulty in maintaining review workflow and audit trail

These challenges significantly increase manual effort and affect reconciliation efficiency in professional practice.

3. Existing Challenges

The traditional reconciliation process using spreadsheet formulas and manual comparison methods becomes inefficient for large GST datasets.

Key challenges include:

• Slow processing for large datasets

• Dependency on VLOOKUP and manual filtering

• Human errors during review

• Performance issues in large Excel workbooks

• Lack of structured exception-based review

• Difficulty in scalable reconciliation processing

The need for a structured, scalable and review-oriented reconciliation platform led to the development of GST Insight.

4. Solution Overview

GST Insight is designed as a scalable reconciliation and audit intelligence platform.

The workflow begins with importing Books data and GST reports such as GSTR-2A and GSTR-2B. The uploaded data is standardized using preprocessing techniques to normalize invoice numbers, GSTIN values, date formats and tax fields.

The standardized data is then loaded into a SQLite database engine. Indexed SQL JOIN operations are used to perform high-speed invoice matching and comparison between Books and GST datasets.

The platform identifies:

• Matched invoices

• Missing invoices

• Duplicate invoices

• Tax differences

• Value mismatches

• Exception-based reconciliation cases

The final results are displayed through an interactive Streamlit dashboard and exported as Excel working papers for professional review.

5. System Architecture

Architecture Flow:

INPUT DATA

(Books / GSTR-2A / GSTR-2B)

       ↓

DATA STANDARDIZATION

       ↓

SQLITE DATABASE ENGINE

       ↓

INDEXED SQL JOIN MATCHING

       ↓

EXCEPTION DETECTION

       ↓

STREAMLIT REVIEW DASHBOARD

       ↓

EXCEL OUTPUT REPORTS

The architecture combines data preprocessing, database-driven reconciliation and dashboard-based review mechanisms for scalable GST reconciliation workflows.

6. Workflow Explanation

Step 1 – Data Upload

The user uploads Books data and GST reports through the application interface.

Step 2 – Data Standardization

The uploaded data is cleaned and standardized through invoice normalization, GSTIN formatting, date conversion and field validation.

Step 3 – SQLite Data Loading

The processed data is loaded into temporary SQLite tables.

Step 4 – Indexed SQL Matching

SQL JOIN operations are performed using key reconciliation fields such as GSTIN, invoice number, taxable value and tax amount.

Step 5 – Exception Detection

The engine identifies missing invoices, duplicate entries, value mismatches and tax differences.

Step 6 – Dashboard Review

The reconciliation summary and detailed reports are displayed through the Streamlit dashboard.

Step 7 – Report Generation

The reconciliation output is exported as Excel working papers for professional review and audit documentation.

7. Technology Stack

The following technologies and libraries were used for development:

• Python – Core application development

• Pandas – Data preprocessing and cleansing

• SQLite – Database-driven reconciliation engine

• SQL JOIN Operations – High-speed invoice matching

• Streamlit – Interactive dashboard and user interface

• OpenPyXL – Excel report generation

The combination of these technologies helped create a scalable and professional reconciliation platform.

8. Key Features

• Intelligent GST reconciliation engine

• SQLite indexed matching logic

• Duplicate invoice detection

• Missing invoice identification

• Tax mismatch detection

• Exception-based review workflow

• Interactive Streamlit dashboard

• Exportable Excel working papers

• Structured reconciliation process

• Scalable processing for large datasets

9. Technical Innovation

The major technical innovation in GST Insight is the transition from traditional dataframe-based reconciliation to indexed SQLite-based processing.

Initially, reconciliation operations were performed using dataframe loops and filtering logic. However, as transaction volumes increased, performance and scalability limitations became evident.

To improve efficiency, the reconciliation engine was redesigned using SQLite and indexed SQL JOIN operations. This architecture significantly improved reconciliation speed, scalability and processing efficiency.

The architecture combines:

• Pandas for flexible data preprocessing

• SQLite for indexed database matching

• SQL JOIN operations for scalable comparison

• Streamlit for interactive review workflow

This creates a structured and scalable reconciliation platform suitable for professional practice.

10. Practical Benefits

GST Insight provides practical benefits for Chartered Accountants and tax professionals.

Key benefits include:

• Reduced manual reconciliation effort

• Faster processing of large datasets

• Improved review efficiency

• Better audit support documentation

• Exception-based review mechanism

• Structured working paper generation

• Professional and scalable workflow

The platform helps professionals focus more on analytical review instead of repetitive manual matching activities.

11. Future Scope

GST Insight has potential for future expansion and integration into broader audit and compliance workflows.

Proposed future enhancements include:

• Tally ODBC integration

• AI-assisted anomaly detection

• Automated notice drafting support

• Cloud-based deployment

• Advanced audit analytics

• AI-powered review workflows

• Intelligent risk scoring mechanisms

• MCP-based AI integration

The modular architecture supports future scalability and technology integration.

12. Conclusion

GST Insight demonstrates how practical automation and scalable data processing can improve traditional GST reconciliation workflows.

By combining Pandas preprocessing, SQLite-based indexed reconciliation and interactive Streamlit dashboards, the platform transforms reconciliation into a faster, scalable and review-driven workflow.

The project highlights the practical application of Python technologies in solving real-world challenges faced by Chartered Accountants and audit professionals.

The objective of the platform is not to replace professional judgement, but to assist professionals by reducing repetitive work, improving reconciliation efficiency and enabling better exception-based review.