DeepJudge has launched AI Workflows, a significant enhancement to its AI-powered search platform designed for internal law firm data. This feature enables law firms to utilize AI agents and applications within their existing data ecosystems, facilitating both pre-built AI workflows and custom applications through a user-friendly low-code/no-code interface.
DeepJudge asserts that this development represents a pivotal advancement in legal AI, offering law firms the ability to craft AI-driven workflows specifically tailored to their internal requirements, all without the need to reformat or externally upload their data. Among the firms capitalizing on this innovation are Gunderson Dettmer and CMS Switzerland. Notably, Gunderson Dettmer, which had gained attention in 2023 for launching ChatGD, the first proprietary internal generative AI app by a U.S. law firm, is now replacing that app with DeepJudge’s platform.
Founded by former Google researchers who hold Ph.Ds. in AI from ETH Zurich, DeepJudge was already recognized for its robust enterprise search capabilities, allowing seamless integration of unstructured documents across a firm’s various data platforms like Document Management Systems (DMS), Microsoft environments, and client portals into a unified search interface. The recently introduced AI Workflows enhance this search platform by allowing firms to develop and deploy applications leveraging their collective knowledge and experience, using either predefined workflows or a no-code builder.
DeepJudge’s CEO, Paulina Grnarova, commented on this approach, emphasizing that their solution negates the need for firms to “clean up” their data, which is a requirement from many other AI vendors. “AI should work for you — not make you work for it,” she stated, pointing out that existing solutions often shift the burden to legal teams who do not operate with perfectly structured information.
The new functionalities include preset workflows that can summarize documents, create overviews by compiling pertinent information, and analyze clause variations across documents. The flexibility extends to complex custom workflows that can be crafted through DeepJudge’s no-code editor.
Yannic Kilcher, CTO of DeepJudge, highlighted the prevalent issue of “data trapped in silos” within legal enterprises. “Much of the work in building AI apps is handling the data integration, data governance and finding the right information for any query,” Kilcher explained, pointing out that legal data often exists in disparate, disconnected systems. DeepJudge’s platform aims to address these challenges by seamlessly integrating with existing systems while respecting current security protocols and access controls.
The AI Workflows feature offers robust search capabilities scalable to millions of documents, complemented by large language models (LLMs) that process the retrieved data. This synergy allows attorneys, for instance, to quickly gather and digest pertinent client history before meetings or rapidly summarize new case contexts, enhancing efficiency and compliance.
Nick Zaugg from CMS Switzerland noted the platform’s value in utilizing existing firm knowledge swiftly and effectively, a crucial aspect for delivering AI-driven client work products. Grnarova shared that the company has been expanding its operations in the U.S., indicating a broadening market presence for DeepJudge.
For further exploration, those interested can join DeepJudge in a live webinar featuring Gunderson Dettmer, scheduled for March 18, 2025.