The context problem: How Docusign built agents that understand the business behind contracts
Ask a general-purpose language model to summarize a contract, and it will probably do a reasonable job. Ask it to The post The context problem: How Docusign built agents that understand the business behind contracts appeared first on The New Stack .
Docusign is developing AI agents that comprehend the broader business context behind contracts, going beyond general-purpose language models. The company's Intelligent Agreement Management (IAM) platform, introduced in April 2024, aims to create, commit to, and manage agreements. A key component of this platform is Docusign Iris, an AI engine that leverages the company's contract management expertise, agreement-centric data, and AI models to deliver insights and enable actions throughout the contract lifecycle.
To differentiate its AI solution from generic language models, Docusign focuses on four categories of context: within-agreement, cross-agreement, organizational, and user. Within-agreement context refers to understanding terms and clauses defined in a single contract. Cross-agreement context involves recognizing relationships between multiple contracts and amendments to determine which version of a term applies.
Organizational context encompasses internal policies, such as legal playbooks outlining acceptable terms, while user context includes factors like language preferences, previous queries, and feedback.
The AI contract agent must assemble and interpret various pieces of information to accurately answer complex questions about contracts. For example, when asked about contracts valued over $1 million due for renewal within the next three months, the agent must identify relevant agreements, account for amendments, and cross-reference them against organizational policies. This process may involve integrating multiple tools and drawing upon Docusign's proprietary data.
While general-purpose LLMs can handle some aspects of within-agreement context, the other three categories pose more significant challenges. Docusign's Iris platform combines intelligent contract extractions, semantic search capabilities, and agreement-specific logic to bridge these gaps. By automating certain contract workflows and enforcing business rules and approval processes, Iris allows organizations to strike a balance between AI-driven efficiency and human oversight.
Written by urgent.news from The New Stack's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
