CEO Josh Baxter notes the system can route tasks to the best-performing model as LLMs evolve.

NetDocuments unveils a reimagined platform built around a first-of-its-kind legal context graph that continuously maps how matters, documents, and communications connect within a firm while preserving permissions and ethical walls.

The platform broadens access to contextual knowledge to improve outcomes across diverse legal workflows, whether using NetDocuments’ AI or external solutions.

The context graph operates on three levels—document, matter, and global—linking and indexing information so AI sees law as a web of relationships, not a simple document collection.

Key user benefits include meaning-based search, preservation of institutional knowledge, rapid matter context at opening, AI-assisted drafting and redlining in familiar tools, and AI agents drawing from a connected, permissioned source of truth.

Questions raised include how supervising partners can audit AI prompts (per-prompt provenance, permissions, model version, retention) and whether the platform provides an auditable prompt trail.

Demo videos and a webinar showcase matter overviews, contextual search, and agentic workflows, with waitlist and enrollment options for firms.

AI agents inside NetDocuments and external tools (via MCP) operate from a context-rich, permission-aware source of truth, reducing duplication and surfacing precedents more accurately.

Features include automatic matter summaries with parties, dates and timelines; natural-language search with source-cited results constrained by access permissions; automatic version-change summaries; real-time data extraction on document adds; and in-Word AI-assisted co-authoring.

The platform also offers cross-matter natural-language search, assembled matter overviews, and an in-Word drafting panel, signaling a broad first-party AI surface area.

For existing NetDocuments customers, the upgrade path is low-friction with no migration, allowing users to switch between old and new interfaces at their own pace.

Built in partnership with AWS and Elastic, the platform is model-agnostic, routing queries to various LLMs and integrating with tools like Anthropic and OpenAI to ensure scalable performance.