The routine task of extracting data from court rulings, contracts, or incoming correspondence takes lawyers hours that could be spent on analysis or client work. Modern technology allows automating this process without requiring programming skills. In this article, we will show how to create your own service for extracting structured data from documents using the GigaChat neural network on the Botman.one platform.
The Problem:
Lawyers daily face the need to extract repetitive data from documents: Tax IDs (INN), company names, case numbers, addresses, amounts, and dates. This is slow, tedious, and prone to errors.
The Solution:
Using the combination of the no-code platform Botman.one and the GigaChat neural network. You create a scenario that:
-
Accepts a document (PDF, Word, image).
-
Sends it to GigaChat for analysis.
-
Receives clearly structured data in the form of a table or JSON.
Step 1: Setting Up the GigaChat Integration
Before creating the bot, you need to connect the GigaChat API.
Important: For correct operation, you must properly configure the request to the neural network so that it returns data in the required format.
🔗 A detailed guide on setting up the GigaChat API can be found here.
Step 2: Creating a Parser Bot on Botman.one
-
Create a new bot and select the "Document Processor" type.
-
Add a file upload block — users will upload documents through the bot interface or API.
-
Configure the GigaChat call via the API block:
-
Specify a prompt (instruction) for the neural network, for example:
"Extract the following data from the document: Tax ID (INN), organization name, court case number, date, address. Return the response in JSON format." -
Pass the content of the uploaded document in the request.
-
-
Process the response — convert the received JSON into a convenient table or send it to your CRM/database.
-
Test the bot on different document types.
Use Case Example:
Task: Automatically extract data from arbitration court rulings.
Solution:
-
The bot accepts a PDF file of the ruling.
-
GigaChat finds and returns:
json
{ "case_number": "А40-123456/2023", "company_name": "OOO 'YUG'", "inn": "7712345678", "address": "Moscow, Primernaya St., 1", "date": "2023-10-26" } -
Botman.one saves this data to a Google Sheet or sends it to your legal system.
Why is This Convenient?
✅ No programming required — the entire process is built with visual blocks.
✅ Flexibility — you can configure the parser for any document type.
✅ Integrations — ready-made connections with Telegram, WhatsApp, Excel, databases, etc.
✅ Speed — from document upload to ready data takes less than a minute.
Conclusion:
No-code platforms like Botman.one blur the lines between law and technology. Now any lawyer can independently create a tool to automate routine tasks, increasing their efficiency manifold.
Ready to automate your document work?
Start with the GigaChat setup guide, and then create your first parser bot on Botman.one.