JSJeong-me/GPT-Table
GPT Table Semantic Parsing with complex & non-intuitive structure.
This tool helps financial analysts and operations professionals accurately extract structured information from complex financial documents like corporate disclosures, invoices, and receipts. It takes a raw financial document, converts its tables into a structured XML format, and then represents the data as easy-to-use tuples for reliable financial analysis and decision-making. The output is a highly accurate, error-free interpretation of tabular data, ready for automated processing or question-answering systems.
No commits in the last 6 months.
Use this if you need to extract precise numerical and categorical data from challenging financial tables, ensuring accuracy and reducing 'hallucinations' when using AI systems for analysis.
Not ideal if your primary need is for unstructured text analysis or general document summarization, as its strength is in detailed tabular data extraction.
Stars
17
Forks
2
Language
Jupyter Notebook
License
Apache-2.0
Category
Last pushed
Jul 16, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/JSJeong-me/GPT-Table"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
gpt-open/rag-gpt
RAG-GPT, leveraging LLM and RAG technology, learns from user-customized knowledge bases to...
LexiestLeszek/scrapeGPT
ScrapeGPT is a RAG-based Telegram bot designed to scrape and analyze websites, then answer...
leon0204/fast-rag
LLM Rag Intelligent Q&A Robot
gptscript-ai/gptparse
Document parser for RAG
maanvithag/thinkai
An LLM app with Retrieval Augmented Generation (RAG) built using OpenAI GPT models, Langchain...