Business Contracts Remain Complex, And Artificial Intelligence Can Improve Contract Management by David A. Teich
Article Highlights | AI and Contracts
- AI as a contract review tool has been used for basic legal discovery for quite some time now, but it is beginning to be applied to the wider issue of full contract management and in a slew of different industries.
- For a more comprehensive analysis that takes into account contract success and failure, teams should examine the data from business documents in combination with ERP, accounting, and other systems.
- Big contract issues are more generic. The larger opportunity for differentiation for each company comes from how they use the data, how they analyze it and then work with their partners and customers to optimize business performance.
- Companies have always struggled with contract management and evaluation. Advances in contract AI software can help us tackle this problem, but this means that current solutions require more than just AI and a bit of programming.
Business contracts are often extremely complex. For instance, utilities infrastructure, whether remote telecommunications towers, pipelines, or major plant construction can cover a massive amount of details, from real estate details, through environmental issues, construction responsibilities and liability, to managing the life of the project. Companies struggle to understand, manage and evaluate the large amount of text involved in business and governmental interactions. While artificial intelligence (AI) has been used for basic legal discovery, it is now beginning to be applied to the wider issue of full contract management.
Companies have always struggled with contract management and evaluation. Large contracts are extremely complex and difficult to evaluate. A large volume of smaller contracts, all with slightly different terms and conditions, all in different parts of their lifecycles, create a different yet still critical challenge.
Document management companies have grown large with basic search of documents, and have added simple business intelligence (BI) analysis. Legal ediscovery systems have been growing for years and providing additional information. What’s needed is a blend of those two and information from ERP, accounting and other systems that have been used to analyze contract success and failure.
“Most industries we’re working with have strong similarities across their contracts with narrow, but important, areas of uniqueness,” said Nick Vandivere. “The larger opportunity for differentiation for each company comes from how they use the data, how they analyze it and then work with their partners and customers to optimize business performance.”
Though contracts have been around for thousands of years, the modern complexity of contracts combines with the young age of AI to mean that current solutions require more than just AI and a bit of programming. In each industry, there are subject matter experts who ThoughtTrace works with to build a model to both understand and to display critical information in an actionable format. That drives the reality that much of the learning in the systems is supervised, with some unsupervised used in a hybrid system to notice unique items and outliers.
As with other companies with whom I’ve recently talked, their machine learning (ML) uses a mix of deep learning and random forest, depending on data volumes and the sparsity of the data in a specific contract area. As should be expected, the bulk of the analysis is language-based, and that includes term sets for different industries.
Information is extracted from the language, and then the contracts are analyzed for metrics, integration with ERP, CRM, and other systems can be leveraged to provide performance analysis. BI analytics and display can help make things clearer to the clients.
One thing to which I responded with vigorous nodding was the answer to my question about how the AI models are delivered to the user. I’ve repeatedly mentioned that AI/ML is another toolset, not a solution. To ThoughtTrace, how the model is developed and what tools are used are things that should be under the hood. Joel Hron, CTO, ThoughtTrace, was very clear that he does his best to obfuscate how AI is operating. If the customer is presented with clear, timely, and actionable information, AI can be mentioned but it isn’t the story. I tell my clients they need to speak to the customer about the customer first, and about their technology second. That’s a story I didn’t have to mention in this case.
Artificial intelligence has had previous winters in its acceptance. Those happened because it, for many reasons, couldn’t do the basics that were promised. Today, there are a whole lot of arenas in which AI can successfully play. The challenge for the current bubble is, quite simply, not to be a bubble. Overpromising AI as a panacea is one way to create a bubble that will burst. Focusing on AI as a critical tool that must be integrated into a system is the way to keep building acceptance. Contracts are one of the most complex things to understand and analyze in the business world. Artificial intelligence can help in contract analysis and management, and is starting to be applied in the real world.
About the Author:
David A. Teich is interested in artificial intelligence (AI), machine learning (ML), robotics, and other advanced technologies focused on how they help businesses improve performance. He’s an analyst and consultant in those areas as well as in high tech, B2B, marketing. Previous work runs the gamut in software, including operations, development, field consulting, sales engineering, and product marketing. He has worked in startups, mid-sized companies, and global organizations.