← Sticker Wall AGENT · DETAIL

Law Firm AI Case Report Generation Pay-Per-Article Service Reaching 5,000 RMB Monthly

Workflow: Receive pending judgment HTML or web links sent by lawyers daily via WeChat and law firm collaboration platforms as inpu

AGENT

Key Fields

FIELD STAMPS

IndustrySaaS / Enterprise Software

RegionChina

ScaleSME

ChannelOnline

🔧 Workflow

Receive pending judgment HTML or web links sent by lawyers daily via WeChat and law firm collaboration platforms as input. Call the open-source legal-case-report tool to parse the page structure, combine it with Tongyi Farui and Claude large models to extract core case facts, disputed focuses, judicial reasoning, and applicable laws, and generate annotated, reviewable Word reports. Deliver to clients after manual accuracy review while retaining the case database to optimize subsequent parsing accuracy.

🛠 Setup Requirements

Setup requires basic Python programming skills. Download the open-source legal-case-report project from GitHub and complete local deployment, configure Claude or Tongyi Farui large model API keys, and possess basic legal knowledge to judge whether key report information is complete. The setup cycle takes about 10-15 days. Initially, ready-made tools can be used to run through the workflow before gradually optimizing parsing rules.

🧰 Toolchain

🔧 n8n
🔧 Claude 3.5 Sonnet
🔧 legal-case-report open-source project
🔧 Tongyi Farui
🔧 PKULAW Database

💰 Revenue

① Pay-per-article judgment summary reports for small and medium-sized law firms (main revenue): Law firms pay per article, 80-300 RMB per article × 60-100 delivered monthly = monthly income of about 4,800-30,000 RMB. Card disclosures indicate monthly revenue of about 5,000 to 12,000 RMB, with about 100% of monthly revenue driven down this path. Numbers are derived from per-article pricing and delivery volume; case study display data has not been independently audited; ② Long-term batch cooperation packages (project service fees/monthly packages): Establish monthly batch cooperation with 3-5 small law firms. The card states revenue can stabilize at the upper limit range; lump-sum quotes are not public, and the proportion of total revenue from the 3-5 partners is not mentioned; ③ Second scenario of pay-per-review: Source material features a similar legal side hustle reviewing 2-3 contracts per workday, 200 RMB × 20 days = 4,000 RMB/month (data provided by case study, not independently verified). This can be replicated as pay-per-call judgment review, though its market share is unspecified; ④ Opportunity: Standardized subscription products for judgment summaries targeting law firms. Source material shows an 83% client renewal rate for similar legal services (also from the case source, lacking independent verification); the potential scale of this revenue remains unquantified.

💸 Cost

Monthly cost is about 300-800 RMB, among which large model API call fees are about 200-500 RMB, PKULAW individual version subscription is about 99 RMB/month, and the remainder consists of minor maintenance costs for document processing tools, which can be diluted to 1-2 RMB per article as order volume grows.

⏱ Time Investment

Invest 3-5 hours daily, of which report generation automation takes about 1 hour/10 articles, and the remaining time is used for client communication, report review, and case database iteration rule optimization.

🚀 Getting Started

Step 1: Download the legal-case-report project on GitHub, complete local deployment according to the documentation, run through test cases, and get familiar with the full process from judgment parsing to report generation. Step 2: Join legal industry communities, connect with independent lawyers and small law firms, provide 3-5 free trial reports to validate effectiveness, and subsequently charge per article or sign monthly batch cooperation agreements.

🔑 Keys to Success

✅ Accuracy of core information extraction from judgments, covering three core dimensions: disputed focuses, judicial reasoning, and applicable laws
✅ Standardization of report format, complying with lawyer reading and court submission standards
✅ Building trust with lawyer clients, accumulating reputation through free trials and after-sales revision services
✅ Continuous iteration of the case database, optimizing parsing rules for common causes of action to improve efficiency

⚠️ 风险

⚠️ Legal documents involve client privacy, and data leakage may lead to serious compliance consequences
⚠️ If the generated report omits key legal details and is not strictly reviewed, it may mislead lawyers
⚠️ Without lawyer qualification, only case information organization services can be provided; legal opinions must not be issued to avoid crossing the red line of illegal practice
⚠️ Format variations among judgments from different regions are significant, requiring continuous updates to parsing rules to prevent drops in accuracy that lead to client churn

📌 Real Cases

📌 DuckDBlab's shared side hustle practice of using Claude to handle legal contract matters, achieving a monthly income of over 5,000 RMB
📌 GitHub open-source project legal-case-report has implemented an automated workflow generating reviewable Word reports and annotated judgments from judgment web pages
📌 Shenzhen Zhiquan Law Firm utilized AI-assisted similar-case retrieval and report generation in 2026, reducing lawyer desk work time by 40% and significantly improving similar-case research efficiency

Sources