Introduction

Analyzing a legal contract is hard work and takes time. For insurers, it is a critical issue. The contract outlines what they will and won’t cover. If the contract behaves unexpectedly, it can lead to legal, financial, and reputational problems.

This is especially challenging for the London insurance market, the global center for complex risk. Think space rockets, cyber attacks, and coffee taster’s tongues. The market comprises about 100 insurers handling approximately 250,000 deals, generating USD 120 billion in premiums annually.

In this market, each deal involves someone who needs insurance, a broker, and several insurers to support the deal. The risk is shared among these insurers because it is too risky for one insurer to handle alone. But there is only one contract. This contract might be prepared by the broker, the person needing insurance, their lawyers, or one of the insurers. Every insurer involved must understand whether the contract meets their requirements.

That’s why this system was created, allowing lawyers and underwriters to upload a contract, receive an analysis, and engage in discussions.

Team size

3 members

Industry

Insurance

Technology

Azure, Azure OpenAI, ChatGPT-4o, React JS, Python

Highlights

Value Delivered

  • Provide analytical tools and detection capabilities for identifying hidden issues within insurance contracts.
  • Facilitate interaction with both individual documents and batches of documents.
  • The Minimum Viable Product (MVP) version has been delivered and is currently undergoing a trial phase with a potential customer in the United Kingdom.

Challenge and Solutions

  • The new technology exhibits variability in accuracy and performance.
  • Significant infrastructure costs are associated with its implementation.