Trellis Expands Agentic Legal Research to ChatGPT and Claude

The integration moves beyond standard chatbots, allowing attorneys to assign research goals that the AI executes autonomously. By connecting to the Trellis dataset, these agents plan strategies, cross-reference verdicts and motions, and verify citations—tasks that previously required hours of manual labor from litigation associates. This architecture addresses the historical fragmentation of state trial court records, which account for 98% of civil litigation in the United States but have long remained difficult to access at scale.

Nicole Clark, CEO of Trellis, distinguishes this approach from models that rely on general memory. By utilizing an agentic system that researches, double-checks sources, and produces structured outputs like memoranda and damage benchmarks, the platform provides evidence-based insights rather than educated guesses. For example, an AM100 partner utilizing the system can generate a detailed analysis of California premises liability cases involving specific retail hazards in seconds.

This expansion is part of a broader strategy to embed litigation intelligence directly into the existing workflows of legal teams. With coverage spanning over 3,500 courts across 45 states, Trellis aims to bridge the gap between appellate law and the practical realities of trial court proceedings, where judicial patterns and opposing counsel strategies are formed. Future updates are expected to extend this agentic access to additional platforms relied upon by legal practitioners.