I’ve seen people use OpenAI’s API to extract data from PDFs or messy HTML.
It’s *kinda* reliable if you chunk the input and chain prompts.
But honestly, unless you’re dealing with unstructured garbage, traditional tools are faster and cheaper.
It’s *kinda* reliable if you chunk the input and chain prompts.
But honestly, unless you’re dealing with unstructured garbage, traditional tools are faster and cheaper.
