E is For Extract
Last week we began to deconstruct the appellation of the market-leading contract discovery and analytics company, Seal Software. We learned that the S in Seal stands for Search – a core part of the technology and an element which no other vendor in the contract analytics space is able to address to such a complete level. We learned the importance of contract discovery and the ability to do simple Boolean-type searches on all contract information. Today we are turning our attention to the E in Seal, which the founders identified as Extract.
Reviewing each contract to find those that contain specific terms or clauses of interest requires knowledgeable legal teams to conduct the review and more often than not the number of contracts that require a review may run into the thousands, and in some cases even millions of documents. Think of a Fortune 100 organization’s procurement department (for example, P&G which has over 75,000 suppliers) with at least one MSA and dozens of related documents. When they need to get an accurate view of some specific contractual data, such as net payment terms or contract renewal dates, across their entire corpus, there really is no way to consistently and accurately extract that data through a human review process. It just isn’t scalable, and it is error-prone. More importantly, are the limitations of the one-time process.
In a manual review, one reads the first document, notes down the details in a spreadsheet and moves to the next one – rinse and repeat a thousand times. Now imagine you finish that exercise and the Chief Security Officer then informs you of a data breach and asks which suppliers and customers must be notified. So, you need to look at the contracts again to determine if there are any clauses in each document that might reveal your obligations in the event of a data breach. Darn it, if only you had known about this sooner you could have extracted that information while reviewing the contracts for the payment terms. Now you have to restart the document review all over again. And therein lies the problem. You don’t know what is around the corner and what data you may need in the future. The complexity and sheer number of times the contracts need to or should be reviewed can quickly consume legal budgets and resources – slowing the business.
That is where extraction technology like Seal can transform the data extraction process from a costly, repetitive, and manual process to an automated AI-assisted process. As we saw in the previous Search post, Seal is able to digitize unreadable contracts and store an indexed version in a repository. During the ingestion process (or on demand) Seal can also extract a large number of key contractual data points. Typically, these are common terms and clauses that a large majority of contracts will contain such as the contracting parties, start and termination dates, payment terms, contract type, change of control, limitation of liability, jurisdiction, renewal date, termination rights and many more. Seal constructs a multi-layered index of the extracted data that will be the basis to support document reviews with standard Boolean and advanced AI-powered search (more on that in next week’s Analyse post.)
However, most of our customers have their own specific extraction requirements that go beyond that standard set of extraction points. These could be industry-specific or use case-specific terms. Then there are the unforeseen events that require looking for un-extracted terms.
Seal recognized this need for use-case specific metadata and metadata generated on the fly and developed data models that address both requirements. This led to the creation of what we call Insight Accelerators which are essentially pre-packaged sets of analytics that have been trained to address a set of business use-cases. Returning to the Fortune 100 example, the full corpus – hundreds of thousands of documents – is subject to both routine and dynamic extraction processes that are consistent, accurate, and significantly faster and more cost-effective than the legacy manual process.
We see a lot of claims by legal AI vendors that they can extract hundreds of data points or clauses or topics. Whilst Seal certainly is able to do many hundreds, the value is in the depth and quality of the extractions – how far into the topic the analytic takes you. For instance, being able to identify a payment clause, either through word matching or meaning is valuable, begets the question, “What is in that payment clause and is it relevant to my search?” You might want to know the late payment fees for example or if there is payment tiering based on the overdue period. You want and need that information surfaced. That’s what Seal’s extractions can do for you.
So, E is for Extraction although it could also stand for efficient, efficacious and effortless!
Next week we’ll take a look at A. Can you guess what that stands for?
To learn more about Contract Discovery and Analytics, download our free eBook.