

An analysis set includes models and analysis profiles.
There are two general categories of models in Contracts:
You can add either category of model to an analysis profile.
Contracts includes the following models that are ready to run:
Contracts model type | Description |
---|---|
Essential Fields | Uses generative AI to auto-populate fields for every document in your data source. These fields include, but aren’t limited to, key data points such as the contract title, contract type, parties, effective date, and governing law. For more information, see Auto-populate essential fields with generative AI. |
Segmentation | Breaks up every document in your data source into sections and creates a new document for every section found. For more information, see Segment contracts into sections. |
Section Classification | Identifies all section types for every document in your data source. For more information, see Classify section types. |
Definition Extraction | Extracts all definitions for every document in your data source. For more information, see Extract definitions from contracts. |
Contract Classification | For regions that do not support the Essential Fields model, we recommend using the Contracts Classification model to auto-populate contract title and type. If you are in a region that uses Essential Fields do not run Classification. For more information, see Classify contract types. |
The following models require customization before running:
Contracts model type | Description |
---|---|
Regular Expression | Runs all selected regex on your data source and either generates a report, auto populates a field of your choosing, or both, depending on your model configuration. For more information, see Run reports on regular expressions and Auto-populate fields with regular expressions. |
Regular Expression Segmentation | Uses a regex you create to break up every document in your data source into sections and creates a new document for every section found. For more information, see Segment contracts into sections. |
Regular Expression Definition Extraction | Uses a regex you create to extract all definitions for every document in your data source. For more information, see Extract definitions from contracts. |
Document Assembly | Generates a new document for every document in your data source based on your assembly template and field-to-token mapping. For more information, see Automate document assembly based on field values. |
Unitization | Auto groups documents into families based on shared field values. For more information, see Auto-group families based on shared field values. |
ML Extraction | Auto-populates text fields by giving you the ability to train a custom machine learning model using previously coded examples as training data. For more information, see Auto-populate fields by training custom ML models. |
The analysis profile includes the following options:
Analysis profile option | Description |
---|---|
Reference Auto linking | Creates links in the document for all instances where the text references a section found through segmentation. |
Batch Size Setting | Determines how many documents an agent will process at a time. If you leave this blank, Contracts will default to the value you have set for the Batch Size instance setting in the Contracts AI Section. |
Notifications | Contracts will send an email to the user selected upon completion of an analysis set. This email will include information on the number of documents completed, canceled, excluded, and errored with a link to your analysis set. |
Following are high-level instructions for running Contracts analysis:
Once you run the analysis, you can review and evaluate your results and rerun analysis as needed.
A common workflow when using Contracts includes running analysis, evaluating results, making changes to your analysis profile, and then rerunning analysis on the same document set to get better results.
To rerun contracts classification:
Note: Contracts will skip any documents for which there is already a value.
To rerun contracts segmentation:
To rerun extraction:
When running Extraction Models, you must check the Model setting for Existing Values. If you populated fields and have the Existing Values setting set to Do Nothing, you must either change it to Replace or Mass Edit to clear out the field value yourself.
If you mass edit the value yourself, this will not remove the green dot next to the field in the Contracts Viewer. To remove this location-based association, use the Clear Send to Field script.
Select your saved search and a field to clear, and Contracts will mass remove the location-based association and the green dot next to the field in the Contracts Viewer.
Why was this not helpful?
Check one that applies.
Thank you for your feedback.
Want to tell us more?
Great!