Skip to content
ViqusViqus
Navigate
Company
Blog
About Us
Contact
System Status
Enter Viqus Hub

Uber's Legal AI Agent Slashes Contract Review Time by 20% Using RAG

Retrieval-Augmented Generation Legal Tech Agentic AI LangGraph Enterprise AI Contract Review
November 16, 2026
Source: InfoQ AI

This summary and analysis were generated by AI from the original article at InfoQ AI and may contain errors (how Viqus works). Read the source for full details.

Viqus Verdict Logo Viqus Verdict Logo 7
Operationalizing AI in High-Stakes Workflows
Media Hype 4/10
Real Impact 7/10

Article Summary

Uber has launched a 'Legal Redlining Agent,' an AI tool integrated into Microsoft Word, designed to streamline contract review by analyzing proposed changes, assessing legal risks, and suggesting policy-aligned responses. The system leverages Retrieval-Augmented Generation (RAG) alongside historical lawyer feedback and hard-coded deterministic rules. The agent interprets external edits, compares them against established legal policies, and suggests acceptance, rejection, or modification. A key architectural improvement involved adding a feedback loop that captures the reasoning behind past lawyer decisions, moving beyond simple acceptance/rejection tracking. The architecture is complex, combining a React/TypeScript front-end with a Python backend orchestrated via LangGraph, using OpenSearch for vectorized rules and Redis for state management. While the agent boosts efficiency, lawyers retain final approval, ensuring human oversight remains central to the process.

Key Points

  • The agent reduces contract review time by over 20% while maintaining 91% accuracy in its recommendations.
  • Its architecture is robust, combining RAG for context, deterministic rules for non-negotiable compliance, and a feedback loop for continuous learning.
  • Lawyers retain ultimate responsibility for all recommendations, ensuring human oversight remains integral to the workflow.

Why It Matters

This represents a significant, practical deployment of advanced LLM agents in a highly regulated, high-stakes domain like corporate law. By successfully integrating RAG with deterministic guardrails and a continuous feedback mechanism, Uber demonstrates a mature pattern for enterprise AI adoption that moves beyond simple chatbot interfaces. It sets a high bar for how industry leaders can operationalize AI to achieve measurable efficiency gains in knowledge work, provided the human-in-the-loop governance is maintained.


You might also be interested in