What Is a Rule?

In terms of the Conga AI Contract Analyzer, a rule is specified criteria in a Blueprint that qualifies matching data for individual Clauses and Fields in a Contract document.

Each Field and Clause in a Blueprint has an associated rule (criteria defined by Conga AI users) dictating whether Contract data, initially identified by the artificial intelligence machine, classifies as Field or Clause data. Rules boost the confidence of the AI machine in the Contract Analyzer to accurately identify Clauses and Fields in a Contract document.

To create rules for a Blueprint, add criteria to Field and Clauses. For more information on adding criteria in a Blueprint, see Conga AI Clauses and Fields and Conga AI Analyze Blueprint Matrix.

How Do Rules Work?

Rules work symbiotically with artificial intelligence in the Contract Analyzer and directly affect the Contract document analysis. A Blueprint that contains well-defined Field and Clause criteria allows the Contract Analyzer to more accurately identify and match Contract data.

During the Contract analysis process, the AI machine first identifies potential matching Clause and Field data by using a pre-programmed base of legal knowledge and user-generated data gathered by machine learning technology. The Contract Analyzer then reviews the rules for each Field and Clause to calculate if the initial identified data fulfills the criteria. The Contract Analyzer automatically selects the Contract data identified by the AI machine's knowledge that also fulfills corresponding rules.

It is important to ensure that Clause and Field criteria is accurate and representative of specific Clause and Field data. Rules comprised of inaccurate criteria or criteria that is too selective can hamper the Contract Analyzer from successfully identifying Clause and Field data. For more detailed information on best practices for rules, see Conga AI Analyzer Best Practices.

Example Clause Rule

In the Lease Agreement Blueprint, there is an Applicable Law Clause that has two Keyword criteria. The criteria specify that the matching clause data must contain both the "governed by" and "laws of" keyword phrases to identify as the Applicable Law Clause.

The Blueprint Analyzer's AI engine initially finds three potential paragraphs in the Contract document that could match the Applicable Law Clause. It then moves on to the rule evaluation (Clause criteria) portion of the analysis. Only one of the three paragraph contains both "governed by" and "laws of" within the text. The Contract Analyzer calculates that two paragraphs do not fulfill the rule criteria and one paragraph does. In turn, the Contract Analyzer identifies the paragraph that fulfills the rule criteria, containing Keyword phrases "governed by" and "laws of", as the Applicable Law Clause.

Example Field Rule

In the Non-Disclosure Blueprint, there is an End Date Field. The criteria specify that the matching field data must be a date value (field type) and have a greater value than the Start Date Field.

The Blueprint Analyzer's AI engine initially finds five date values in the Contract document. The Contract Analyzer uses the aforementioned criteria to evaluate if any of the five date fulfills the rule. Only one matched field value is greater than the Start Date Field's value and fulfills the field criteria. The Contract analyzer selects this date value based on the rule criteria fulfillment.