Bridging Internal Knowledge and External Authority

Thomson Reuters (TR) has deepened its partnership with DeepJudge to create a unified workflow between a firm’s internal document management system (DMS) and TR’s AI-powered research and drafting tools. The integration allows legal professionals to search internal firm knowledge—such as past agreements or litigation filings—and immediately feed that data into CoCounsel workflows without manual downloading or re-uploading. This enables users to combine internal precedent with external guidance from Practical Law and Westlaw, ensuring that work product is grounded in both firm-specific expertise and authoritative legal research.

Technical Implementation and Security

The integration is designed to be seamless, utilizing existing user credentials and maintaining established ethical walls and document permissions. By embedding DeepJudge’s search directly into the CoCounsel web application, Westlaw Advantage, and the Practical Law Dynamic Tool Set, TR aims to reduce friction in the research process. Future development plans include deeper integration via the Model Context Protocol (MCP), which is intended to further enhance the interoperability of these systems while maintaining the core requirements of trust, transparency, and verifiability.

Balancing Token Efficiency with Output Quality

As agentic AI workflows increase token consumption across the industry, TR is focusing on outcome quality rather than simple cost minimization. According to CTO Joel Hron, the strategy is to optimize the context provided to the AI. By using DeepJudge to surface highly relevant organizational knowledge, the system can ground its reasoning more effectively, which improves accuracy and ensures that token usage is directed toward high-quality, reliable results that lawyers can stand behind.