A fundamental process for making any AI effective is to train the system with examples and for it to learn to extract patterns and constructs from documents it has yet to see. That, in essence, is what “learning” in this context is all about – learning from examples and then continually refining to improve precision and recall. What sets the Seal AI platform apart (among other things) is its ability to actively learn.
Over the past few weeks, we’ve taken a look at the origins of the Seal brand name. It does not refer to an aquatic animal nor the method of affixing a stamp to legal records, instead, it describes the core functions of our platform. We outlined Search, Extract and Analyze in previous posts and now we are on to ‘L’ as we round out and bring closure to the series. The ‘L’ in Seal was initially designated to stand for “learn”.
The learning process provides end-users with a self-serve ability to add an unlimited number of additional custom phrases/clauses into the extraction analysis process – building a metadata repository that uniquely matches their needs. Organizations can address multiple identified market areas of contract risk as well as quickly address any future areas of risk as they are identified. Additionally, as the organization provides more example language to Seal, the system learns, and the extraction process is improved across the body of contracts. This process, User Driven Machine Learning (UDML), enables sophisticated Machine Learning (ML) technology to optimize the AI-model for very complex data extractions, whilst keeping it user-friendly and simple to use. For an in-depth view of the learning process, read this post from our Machine Learning team.
Earlier on, I said that the ‘L’ initially stood for ‘learn’, and today it still does, as this is at the heart of the Seal platform. However, today the ‘L’ also stands for “leverage”. When we talk about leveraging Seal, we are talking about how the insight extracted from contracts can be surfaced and applied in a variety of ways. Contract specific information such as contract type, formats, terms, and renewal information is delivered via standard, out-of-the-box analytic reporting, and dashboards. It can also be delivered via integration with other corporate legal and data platforms like Relativity, DocuSign, Coupa, and SAP Ariba to provide more connectivity and visibility, enabling intelligent business decisions based on combined sets of data.
Seal analytics provides you with insights to make innovative choices that lead to ROI and positive impact. Those decisions are based on the full visibility of data that is generally inaccessible due to the unstructured nature of important contractual obligations. An example might include tying contractual information with the spend data in a spend management system in order to identify whether payment terms are being observed, missed or made too early. Understanding these terms based on what was contractually agreed to, allows users to gain actionable insights such as whether or not working capital being unnecessarily tied up.
And, there we have it. With ‘L’, we have rounded out our four-part series looking at the origins of the Seal brand name and core product. As the old Ronseal advert used to say, “It does what it says on the tin,” and in our case, we do what we say in our name!
To learn more about Seal and the Contract Discovery and Analytics market, download our free eBook.