Automated LLM Benchmarking on Legal Documents via Google Sheets & OpenRouter

Overview

This workflow automates the benchmarking of large language models (LLMs) on legal documents by integrating Google Sheets, Google Drive, and OpenRouter within n8n. It streamlines the process of uploading, extracting, and evaluating legal documents, providing structured performance insights.

Key Features

• Automated Triggering: Initiates via manual trigger or webhook for flexible execution.
• Document Handling: Retrieves legal documents from Google Drive and extracts content, supporting PDF detection and processing.
• Data Management: Fetches test cases and stores results in Google Sheets for easy tracking and analysis.
• LLM Evaluation: Utilizes OpenRouter and LangChain nodes to process documents and parse structured outputs.
• Conditional Logic: Employs IF and Merge nodes to handle different document types and workflow branches.

Benefits

• Efficiency: Eliminates manual data entry and document handling, saving significant time.
• Accuracy: Ensures consistent benchmarking with automated, repeatable processes.
• Centralized Data: Consolidates results in Google Sheets for transparent performance tracking.

Use Cases

• Legal teams benchmarking AI models for document review.
• Research teams evaluating LLMs on compliance or contract analysis.
• Organizations seeking to automate and standardize AI performance testing on legal data.