AI makes contracts richly searchable
When Microsoft introduced the cognitive search feature of Azure Search in May 2018, Icertis recognized it as a game-changing technology for contract management. “Basic keyword or metadata search is essential and an important feature in contract management,” says Monish Darda, Chief Technology Officer and Cofounder of Icertis. “But if you can use AI to identify and index cognitive information in contracts and use that data to enrich searches, contracts become much more valuable.”
In the past, companies that wanted to apply AI to content within contracts would have to painfully stitch together a series of models across vision, language, and speech to fully extract meaning from their data. Orchestrating and maintaining a series of AI models could then easily become a burden and full-time role for development teams. With the cognitive search pipeline in Azure Search, developers can easily apply a series of prebuilt and custom cognitive skills to enrich and bring structure to their data before it gets indexed. “We fell in love with Azure Search because of its AI skills pipeline,” Darda says. “We can plug in the appropriate cognitive skills based on what the customers want to discover.”
Cognitive search in Azure Search makes it easy to extract insights with AI because of its built-in cognitive skills. Icertis started by integrating three built-in cognitive search skills into ICM:
On top of the built-in cognitive skills, Icertis added custom skills to enhance Azure Search with the ability to understand the legal context of contracts. With these skills, ICM knows how to parse contracts, break them into clauses, and classify those clauses. It also knows how to discover obligations so that companies can avoid, for example, the penalties associated with a failure to fulfill contract terms.
When a contract in ICM is processed by the cognitive search pipeline and the results are indexed, Icertis customers can ask questions against the contract. For example, salespeople drafting a new contract can search all existing contracts with customers of similar size, industry, or payment terms and analyze the contract terms. They can identify risks that made other contracts fail and see revenue opportunities attached to the conditions, such as restricting rights to certain countries. With these insights, negotiators can make the contract more likely to maximize revenue and close it faster.
Cognitive search is even more valuable in uncovering more nuanced content—words and phrases that are related to the search topic but are not exact keyword matches. For example, a data security officer might want to see all contract clauses that pertain to the European Union’s General Data Protection Regulation (GDPR). Cognitive search (because of the ICM custom skills) knows that GDPR pertains to privacy and knows what privacy clauses look like, so it can find those “needles in a haystack” that would never be found with a traditional keyword search.
“It would be humanly impossible to analyze hundreds or thousands of contracts to reveal complex patterns,” Darda says. “But with AI, companies can now search for those patterns quickly and derive the benefits of good contract management.”
After the contracts are signed, cognitive-search-enriched ICM helps ensure that the contract terms are enforced. For example, businesses deliver the products or services that meet the customer’s expectations and get paid. If the customer can correlate the transactions to the contracts and validate what is happening in the environment surrounding the contracts, it can identify when business rules are violated, identify which suppliers always meet their commitments, and reach out to a customer with a specific solution.