Hello everybody,
I’m looking for advice to correct environmental PDF automatically with langchain/graph. If you have any recommendation, or if I’m totally off, please don’t hesitate! or source/cookbooks for help.
More details:
- Goal of this is to automatically correct what is wrong in the pdf, based on the documentation in the vector database
- Vector database is from around 50 PDF with laws about environment that contractors need to respect
- INPUT is PDF is a report after a job. OUTPUT the corrected PDF, and a PDF listing everything that changes with the citations and the source
- each rectangular object here is an agent
- I’m not sure how the agent will work yet
Here is the flowchart, any questions let me know!
