How a Lawyer Uses AI to Run a Complete Litigation Workflow

Overview

Lawyer Cheng Jiandu of Beijing Hairun Tianrui (Zhengzhou) Law Firm walks through a full litigation workflow for an equipment‑sale dispute using the WorkBuddy AI workspace. Instead of showcasing flashy AI outputs, he focuses on how AI can assist each step while the lawyer remains in control.

01 Context – From Conversation to Workspace

He likens case materials to an iceberg: the visible surface includes contracts, invoices, chat records, photos, and trial transcripts, while beneath lie oral statements, negotiation details, commercial goals, acknowledged facts, and unverified judgments. Scattering these materials across separate chat sessions makes it hard for AI to reference the correct source, so he consolidates everything in a single workspace.

02 Running the Full Process

The case is broken into seven stages: material intake, pre‑litigation analysis, party verification, legal source verification, trial preparation, post‑trial handling, and final document generation. AI reads the documents, reports what it has seen, and flags missing or conflicting information before any analysis is performed.

He instructs the AI to first “inventory the materials, don’t analyze.” The AI quickly produces a material summary, but Cheng treats this as raw data, not a finished analysis.

03 Tool Invocation and Professional Judgment

After establishing the facts, the AI queries an enterprise‑information tool to retrieve company details, key personnel, related entities, and litigation history. Cheng evaluates the tool’s output, ensuring the AI’s suggestions (e.g., whether to secure assets) are weighed against client goals and case risk before a final decision is made.

04 Multi‑Agent Collaboration

Complex tasks are divided among different agents: evidence sorting, timeline construction, legal research, draft drafting, and adversarial‑view checking. Each agent works in parallel within the unified workspace, mirroring a real lawyer team where the lead attorney makes the final judgment.

05 Legal Source Verification

AI may hallucinate outdated statutes or nonexistent cases. Cheng therefore mandates that every legal citation be traced back to the original source, confirming name, text, and validity, and that any unverified content be clearly marked as such.

06 From Trial Prep to Document Output

During trial preparation, AI helps outline evidence, identify dispute points, and simulate opposing arguments. It also drafts pleadings, which the lawyer refines, ensuring each step leaves a traceable artifact—material lists, fact matrices, legal bases, risk notes, version history, and final documents.

07 Judgment Value Increases with AI Strength

Cheng concludes that the more capable the AI, the more valuable the lawyer’s judgment becomes. Lawyers must still verify facts, validate legal sources, and decide on strategy, while also managing data‑security boundaries for sensitive client information.

08 Turning the Process into a Reusable Skill

After completing the case, Cheng asks the AI to summarize the workflow into a reusable “Skill” that encapsulates material intake, legal verification, evidence checks, document templates, and delivery requirements. Future similar cases can start from this validated skill, avoiding the need to rebuild the process from scratch.