Bulk Extraction with Contracts AI

Process documents directly from external repositories without uploading each file individually. Use Data Cloud and Document AI to extract large volumes of contract data and convert unstructured data into structured records. Create templates to define which contract fields and clauses to extract.

Documents stored in sources such as SharePoint, Amazon S3, or Google Cloud Storage are brought into an Unstructured Data Model Object (UDMO), which acts as the source for extraction. Extraction runs asynchronously. After you create and submit a configuration, Document AI processes documents in batches and writes extracted data to output Data Lake Objects (DLOs).

1. Set Up Bulk Contract Extraction
   Bulk contract extraction uses Einstein Generative AI, Document AI, and Data Cloud.
2. Connect an External Repository for Bulk Contract Extraction
   Connect an external document repository to Data Cloud so that bulk extraction can access your contract files without uploading them to Salesforce storage.
3. Create a Bulk Extraction Configuration
   Define which documents to process in your connected external repository, the data to extract, and where to store the results.
4. Verify Bulk Extraction Results and Create a Contract Record
   Before creating contract records, validate the extracted data, such as contract details and clauses.
5. Change Ownership of Bulk Extraction Results
   As a Salesforce admin managing extractions for a team, assign extraction results to the appropriate users for review and contract creation by changing ownership.
6. Delete a Bulk Extraction Configuration
   To delete a bulk extraction configuration, you must first delete all associated contract extraction results. Deleting a configuration doesn’t remove previously generated extraction results.
7. Considerations for Bulk Extraction
   Understand the limits, permissions, and supported configurations for bulk extraction.