THE PROBLEM

GST practitioners struggle with: (1) Tracking 250+ circulars across 32 website pages manually, (2) Time-consuming search through multiple PDFs, (3) AI hallucination risk with general-purpose tools, (4) Missing critical updates affecting client advisory. Professionals many times rely on informal groups for such information to get the same quickly

THE SOLUTION

Two-component system: (1) Three Python scripts automate downloading, metadata extraction, and organization of all GST circulars and notifications from gstcouncil.gov.in into year-based folders with standardized naming, (2) Upload organized circulars to Google NotebookLM to create a private, hallucination-free AI knowledge base with source attribution.

HOW IT WORKS

① Python Script 1 downloads PDFs from GST Council website → ② Script 2 extracts metadata to Excel (Circular No., Date, Subject, URL) → ③ Script 3 organizes files into year folders with naming pattern and also creates a single files for circulars and notifications → ④ Upload to NotebookLM → ⑤ AI-powered search with zero hallucination and source attribution

KEY BENEFITS

Time Savings ✓ Accuracy Coverage Cost

85-90% faster research Zero hallucination 100% of circulars Free tools

TECHNOLOGIES USED

Python 3.8+ | Requests & BeautifulSoup (web scraping) | Pandas & OpenPyXL (Excel) | Google NotebookLM (AI knowledge base) | Regular Expressions

IMPLEMENTATION

Setup Time: 20-30 minutes one-time | Prerequisites: Python 3.8+, Google account

MEASURABLE IMPACT

Reduces manual research from 2-3 hours to seconds • Provides instant answers with source attribution • Eliminates risk of missing critical updates • Enables faster, more accurate client advisory • Scalable Income Tax or any other laws for which the data is freely available