Document Q&A leveraging LLMs and Clarifai

Notes

Module Document Q&A

This is a demo app showcasing the usefullness of Retrieval Augmented Text Generation. This module includes 4 main pages:

  1. Upload: This page takes a pdf, parses it into chunks, and uploads it to a Clarifai app with metadata to track the source and page/chunk number.
  2. Upload with Geo: This page takes a pdf, parses it into chunks, then extracts relevant locations using a LLM and uploads the text chunks and geo data to a Clarifai app.
  3. Investigate: This page showcases different use cases such as semantic search for document retrieval, followed with NER (named-entity recognition), document summarization, and a "chat with the document" experience.
  4. Geo Search: This page uses Clarifai semantic search and geo data search to retrieve relevant documents within a designated geo location, which can then used to ground the LLM generation for a summarization task.

Install this module into your app to upload the documents and analyse them!

Here's a blog you can refer to https://www.clarifai.com/blog/doc-qa-revolutionizing-document-analysis

Screenshots

Upload PDF

image

Upload with Geo

image

Investigate

image

Named-Entity Recognition

image

Document Summarization

image

Chat with the Document

image

image

  • Module ID
    docQA
  • Latest Version ID
    0_0_10
  • Description
    Document Q&A leveraging LLMs and Clarifai
  • Last Updated
    Aug 31, 2023
  • Repository
  • Commit
  • Share
  • Badge
    68796