Intraspexion Reviews

Intraspexion Description

The annual expenses associated with commercial tort litigation targeting businesses, which encompass benefits paid, losses, legal fees, and administrative expenditures such as document collection and attorney meetings, were estimated at around $160 billion, culminating in nearly $1.6 trillion over the course of a decade. To create a deep learning model focused on the federal court’s civil rights-employment category, specifically "employment discrimination," we gathered factual allegations from 400 federal court complaints—excluding any email data. This model was deployed on GPU instances within Microsoft Azure and AWS, where it was evaluated using 20,401 emails from the Enron dataset, marking the first instance the model encountered email data. Each identified true positive email can be exported to a platform for internal investigations or case management purposes, enhancing the model’s utility. Furthermore, with an integrated database connected to the user interface, users have the capability to save these true positives, incorporate them into the initial training set, and subsequently re-train the model for improved accuracy and performance. As a result, this continual learning process ensures that the model evolves and adapts over time to better identify relevant cases.