I want to extract text from uploaded PDF documents (uploaded as images) How would I do this?

CrewAI Community Support
1

Hi,

I have uploaded PDF documents where the text is uploaded like images. I’d like to build a RAG tool to be able to extract information from these PDF documents. Which tool should I use to do this?

2

if it’s an online pdf file you could use
[CrewDoclingSource](Knowledge - CrewAI

3

Have you tried Multimodal agents? Using Multimodal Agents - CrewAI

4

No, this is not multimodal use case. My use case is simple to read file from S3 bucket and then extract key fields from PDF in semantic or fuzzy search in pdf. I wanted to understand about PDFSearchTool. How it works ? Do we need to stitch any vector database to perform semantic search or any default support available in crew ai