Not sure where to start? We’ve curated the best legacy contract migration/data extraction practices to get you started
It’s important to cover the best practices, partially covered in the previous blogs.
1. Decide which contracts you want to migrate
2. Decide what data points you want to report on across a set of these legacy contracts
3. Set up rules for extraction for each of the data points, including exception handling
4. Configure your AI solution to extract these data points
5. Configure your CLM for ingestion of these attributes with the same rules.
6. Test ingest smaller batches to ensure smooth bulk ingestion
7. Ensure your AI software forces a legally trained person to visually check and correct issues with data points in the original document (smudgy scans, missing data points, conflicting information, etc.)
8. Ensure the software can handle the bulk of files -with automatic file classification, OCR, matchups of masters and amendments, creation of unique ids, detecting unsigned, partially signed, and irrelevant documents
9. Ensure AI software can add a human check to each of these processes above
10. Software should also be able to batch out the document so multiple people and collate the extraction results when the q/c is done.
11. Better AI software allows for multiple checks by independent legal personnel.
12. If you have paper documents, scan them independently (Masters in one file, addendums in another file) and not group multiple documents into one file. Scan at a minimum of 300 dpi. You can choose OCR as an option, or most providers will OCR the file for you. OCR should be done “image over text”.