Smart Document Assistant for Legal™ (SDA for Legal) is an artificial intelligence application that automatically associates an incoming document with the appropriate legal matter at a law firm or legal department, classifies its type, titles the file, and stores it in the proper folder within iManage®.
Designed to streamline the manual document profiling tasks necessary for all incoming documents, the software leverages machine learning to continuously improve on its abilities.
The software even learns over time whom to notify about a document’s arrival, and facilitates those emails.
Smart Document Assistant’s goal is to facilitate the fastest, easiest, and most accurate ingestion of documents into iManage.
Natural Language Processing technology - enables SDA to profile unstructured legal documents
- Facilitates the processing of the lengthy, all-text documents that can dominate law office files
- Identifies and utilizes all of a document’s informational words and phrases:
- To classify specific document types even if the differences are subtle, without having to rely on spotting a specific keyword
- To help find the normalized client name used by the firm to file their documents
- To enable discovery and association of the correct internal legal matter tracking number
Machine Learning technology - enables constant accuracy improvement
- The software learns:
- To recognize and classify specific document types
- To associate specific combinations of document data with specific matters
- To suggest specific people to send email notices of a new document’s arrival by matter number and user
- The software improves upon what it has learned whenever a user accepts or corrects its recommendations:
- To improve its recommendation accuracy
- To learn how to classify new, previously unseen document types, and build upon its pre-trained knowledge on a variety of legal litigation/court documents
- To learn the criteria for associating documents to each matter number, including just-added matter numbers
- To learn each user’s preferences on who they send new document notifications to - by mater number
Predictive Analytics technology - enables best-estimate suggestions
- The software calculates a confidence percentage on the accuracy of each recommendation:
- To allow reasonable predictions when there is uncertainty instead of being limited to infrequent “perfect matches” with previously seen training documents
- To provide color-coded predictions that call attention to uncertain suggestions
- To maximize machine learning feedback opportunities to improve accuracy
Workflow Automation technology - ensures optimal user productivity
- File names that comply with each firm’s naming conventions can be automatically assembled and suggested using a configurable rule
- Documents are submitted to the appropriate folder in iManage after a profile is verified by a user
- Email notifications with links to the document in iManage can be automatically sent to the suggested list of people, and optionally, to the docketing team from within SDA’s user interface
Confidentiality controls – help keep sensitive information from being revealed
- Software is deployed on premise behind a corporate firewall or in a private cloud
- Authenticated user login and access permissions are synchronized with Active Directory or with iManage's access permissions
Subscription licensing model based on the total number of pages processed in a year. Just three volume/price levels, with unlimited users, on-premise or in private cloud.
Any law firm or corporate law department using iManage. Solution was pre-trained on thousands of public litigation documents from the US court system
iManage 8.5 or later