In the rapidly evolving landscape of legal technology, Knowable has unveiled a suite of generative AI-powered tools, dubbed Ask Knowable, designed to redefine how legal teams manage and interact with contracts. The new tools will launch for general availability in March 2025, but have been made available as a commercial preview as of today. This initiative significantly leverages large language models (LLMs) to tackle the intricate challenges of contract management.
According to Knowable’s CEO Nik Reed, the objective of the suite is to make contract data work for users. By harnessing both new and traditional AI techniques, the platform provides a unique approach that addresses the complexities inherent to managing executed agreements. This involves a blend of generative AI’s conversational capabilities, traditional machine learning, and meticulous human-in-the-loop quality control processes.
Key features in the Ask Knowable toolkit include:
- AI Contract Search: This allows users to locate contracts in seconds through conversational prompts, eliminating the need for cumbersome filters or specialized data modeling skills.
- AI Contract Summaries: Users receive automatic summaries of individual contracts as well as comprehensive overviews of entire contract families, indicating which terms are currently controlling.
- AI Chat: Provides insights into specific clauses and key terms, enabling users to track the evolution of contract terms across interconnected agreements.
The effectiveness of Ask Knowable’s functionalities hinges on two core principles: accurate data extraction, which promises a minimum 98% data accuracy, and the importance of family mapping. These principles are pivotal given the inherent challenges of poorly organized contract data, which, as Reed illustrated, can lead to costly errors across organizations managing extensive contract portfolios. The company cites a case where its technology identified a significant billing error, resulting in considerable savings for its client.
During a demonstration, Reed elaborated on the platform’s ability to manage complex agreements, emphasizing that true contract understanding requires robust family mapping and accurate data. These features allow users to see how contract terms change over time and interrelate among various agreements.
Moreover, to ensure accurate AI responses, Knowable utilizes a Closed Data Environment for its LLMs, which operate solely within the client’s contract repository. This setup minimizes the potential for irrelevant or erroneous outputs. Furthermore, expert-driven prompt engineering is employed to fine-tune the AI’s reasoning processes.
Ask Knowable is anticipated to enter alpha testing with select clients shortly after Thanksgiving 2024, with a broader beta release and subsequent general availability aligned with the Legalweek conference in early 2025.
For more detailed insights into the launch and capabilities of the Ask Knowable suite, you can visit the LawNext article that provides further context on Knowable’s latest developments.