In a move that marries a massive database of state litigation data with the generative capabilities of artificial intelligence, Trellis, a legal technology company known for its analytics tools and extensive collection of state trial court records, today released Trellis AI, a new set of generative AI tools aimed at simplifying litigation tasks. The tools […]
In a move that marries a massive database of state litigation data with the generative capabilities of artificial intelligence, Trellis, a legal technology company known for its analytics tools and extensive collection of state trial court records, today released Trellis AI, a new set of generative AI tools aimed at simplifying litigation tasks.
The tools are designed to help trial court litigators save time and enhance productivity on common tasks such as motion drafting, case analysis, and discovery by automating much of the manual work these tasks would typically involve.
In a demonstration of these new AI features, founder and CEO Nicole Clark emphasized that they build on Trellis’ core strength, which is its database of millions of state trial court motions, rulings, and filings. The company uses this data to train AI models to deliver results that are tailored to specific jurisdictions and case types, resulting in outputs that promise to be more precise and relevant.
The announcement represents a notable expansion for the company, as it moves beyond data search and analytics into the realm of AI-driven litigation support.
“What we’re doing with Trellis AI is we have basically a legal productivity platform that is built on top of all this data as a foundation,” Clark said. “And we’re building this specifically for trial court litigators, so this is using the court system, using the data foundation that we have, to build products and output that litigators can actually use.”
Motion Drafting
The AI tools released today fall under two types, Clark said: on-case tools that generate results related to the case as a whole, and on-document tools that perform functions specific to a given litigation document.
Clark says she is most excited about the on-case tools, and one of the most notable of those, which Trellis calls Draft Arguments, is an argument drafting tool that enables lawyers to use generative AI to draft motions – for now, motions for summary judgment and motions to dismiss, with others to be added over time.
Drawing on successful arguments from other cases, Trellis AI can help draft arguments for a motion, showing the sources it used so lawyers can verify the recommendations.
Drawing on Trellis’ database of litigation documents, the tool uses AI to identify successful arguments from cases that are factually and legally similar. It then generates a draft motion tailoring those successful arguments to the facts of the user’s case.
To create a draft, an attorney starts with one or more case documents, such as a complaint or discovery responses, either already on Trellis or that the attorney uploads, and the tool will search Trellis’ database for cases with similar facts. It then analyzes the arguments used successfully in those cases, applies them to the uploaded materials, and creates a draft motion.
The draft motion is structured in a standard legal format and includes citations to relevant case law and statutes. Each argument is linked to its source document so that the attorney can view the document and verify the output.
As of today’s launch, this tool is available only for California cases, although Trellis plans to expand it to other jurisdictions.
Case Assessment
Another of the on-case tools released today is a case assessment feature that provides a detailed overview of a case, producing a case report that combines factual, legal, and strategic analysis, including key facts, claims, defenses, potential outcomes, and recommended next steps.
Trellis AI’s case assessment tool provides a comprehensive overview of key facets of a case.
It is intended to serve both as a starting point for litigators at the beginning of a case and as a playbook and reference source as they progress through the stages of the case.
Here again, the user would start with the complaint or any additional case documents, and Trellis AI will generate a report that includes:
- Summary of key facts and legal issues.
- Analysis of the causes of action, the evidence required to prove them, and gaps in the evidence.
- Recommended next steps, including potential defenses, procedural tactics and discovery.
- Witness preparation.
- Jury analysis, including demographic and verdict trends in the relevant jurisdiction.
Clark anticipates that this document is one that attorneys will continue to return to throughout the lifespan of a case. As new documents, such as discovery responses or deposition transcripts, are added as the case progresses, the overview can be refreshed and regenerated to reflect the additional information.
This feature is designed to help attorneys strategize about their cases while also keeping track of immediate deadlines and tasks.
Document Analysis Tools
The complaint tool can recommend defenses to the causes of action alleged in a complaint.
Trellis AI also includes a set of document-level tools for analyzing complaints and other case documents. These include:
- Complaint Tools. These tools focus on the complaint, analyzing it and outlining the causes of action, key facts and relief sought, while also suggesting defenses to the causes of action and the statutes of limitations applicable to each.
- Document Analysis. This tool analyzes the arguments in a document, such as a motion, breaking down essential elements of the arguments and defenses, key evidence, gaps in evidence, key witnesses, and more.
- Review Tools. These tools can be used to create summaries of case documents, with options for brief or detailed outputs, and chronological timelines.
- Research/Writing Tools. These tools can extract citations for documents and key terms for conducting legal research. It also includes a tool for generating an email to update the client on developments in the case.
These tools can be used with documents hosted on Trellis or uploaded directly by the user, offering flexibility for attorneys working with external case materials.
‘Hundreds of Millions’ of Documents
Trellis differentiates its AI from other legal AI tools by the fact that it leverages its proprietary database of trial court documents. This allows the company to train AI models that are specific to state laws and procedural rules, reducing errors that can occur when using broader, less focused datasets.
Trellis AI can evaluate the evidence in a complaint and suggest actions for opposing counsel.
“Trellis is uniquely positioned to support attorneys within the very court system where their cases are filed.” Clark said. “Trellis provides access to hundreds of millions of motions and briefs, representing millions of hours of attorney blood, sweat, and tears in research and drafting.”
All of this happens with workspaces that attorneys can create for each case. Users can create workspaces either from docket documents already on Trellis or by uploading documents of their own.
To ensure transparency and to enable users to verify results, each AI-generated output includes citations and links to the source documents.
Trellis AI is designed to enable lawyers to use its features at the click of a button, avoiding chat-based interfaces that require users to become prompt engineers.
Trellis says that it adheres to strict security protocols. All uploaded documents are encrypted at rest and align with SOC 2 standards, it says.
Bottom Line
At launch, the motion drafting tool is limited to California cases. Other features, such as document analysis and discovery tools, are not specific to any jurisdiction. Trellis plans to expand its motion drafting capabilities to other jurisdictions, starting with New York, Texas and Florida, in the near future.
Clark said that feedback from beta users has been positive, particularly regarding the motion drafting and case assessment tools. Early adopters have highlighted the potential time savings and the ability to focus more on strategy by automating routine tasks, she said.
I have not directly used any of these tools, but judging by the demonstration Clark gave me, I was impressed with what I saw. Particularly impressive were the on-case tools that use generative AI and Trellis’ own extensive collection of litigation data to suggest arguments and strategies.
To be clear, there are other AI-driven brief-drafting tools on the market. But the fact that Trellis AI is drawing from an extensive database of court filings seems to have the potential to be particularly powerful.
I was also impressed with the case report feature, which gets quite detailed in providing analysis of causes of action and defenses, facts and witnesses, timelines, tactics, discovery prep, and much more.
We should all know the mantra by now that AI is no substitute for a lawyer’s judgment, but it can certainly be a powerful tool for bringing together and organizing the information that supports that judgment. That, it seems, is what Trellis AI does best.
