At Gallagher Bassett, we're committed to investing in technology that helps our clients work smarter, simplify complex processes, and make more informed decisions.

We're excited to introduce GB AI Assistant, our latest generative AI capability within Luminos, Gallagher Bassett's risk management information system (RMIS). GB AI Assistant allows you and your teams to ask questions about a claim in your own words — and get a quick reply without having to search through the claim details or running reports.

See GB AI Assistant in Action

From questions to decisions, faster

Finding a specific piece of information within a complex claim can take valuable time. GB AI Assistant is designed to change that.

Whether you need to understand what happened following an incident, check the progress of a claim, or quickly get up to speed on a complex defense strategy, GB AI Assistant helps you find the answers.

And because every answer includes clickable source citations, users can quickly trace information back to its original source for greater confidence and transparency.

Key features of GB AI Assistant

  • Natural language claim facts: Simply ask questions in your own words — avoid hunting through notes and claim pages to find the facts you are looking for.
  • Smart summaries and explanations: Provides quick answers to your claim-related questions.
  • Clickable source citations: Instantly trace each fact to its original source for fast, confident validation.
  • Voice input mode: Easily ask your questions by speaking instead of typing — perfect for when you're on the go with your mobile device.

See what real clients have to say about GB AI Assistant

Helping claim professionals focus on what matters

Instead of spending valuable time working through claim files, users can ask questions such as:

Learn more about the technology behind the GB AI Assistant

Technical capability Client benefit How it works Technology under the hood
Secure LLM Access and Monitoring Clients get fast answers to their questions without sacrificing strong controls for privacy, governance, and business risk.
  • Client data stays within GB-controlled infrastructure and is processed through a GB Azure tenant
  • Data is used to generate responses but never used to train foundation models
  • Prompt firewalls and guardrails help reduce unsafe, biased, or unauthorized responses and block malicious and unsafe prompts 
  • Continuous monitoring helps detect issues before they affect users 
  • Azure AI Foundry-hosted LLM accessed through Azure APIs 
  • Azure Application Insights 
  • Prompt firewall 
  • Internal guardrails 
  • Azure AI Content Safety 
  • TLS encryption in transit and encryption at rest 
Agentic RAG for Real-Time Retrieval Clients receive answers that are specific to the question, current to the claim file, and traceable to source evidence.
  • Purpose-built agents retrieve relevant information from the claim file
  • A synthesis agent creates an answer grounded in the retrieved evidence, with citations where available
  • Agentic RAG pipeline 
  • LangGraph multi-agent framework 
  • Hybrid search using BM25 and vector search, with semantic re-ranking 
  • Dynamic top-k retrieval with date-filter fallback 
  • Chunk relevance scoring 
  • Citation-backed synthesis 
  • DeepEval automated LLM evaluation 
API-First Design Clients can use the same Q&A capability inside their own applications, not just through the chatbot interface.
  • Secure APIs provide access to the same multi-agent RAG pipeline used by the chatbot
  • Prompt firewall controls help protect against unsafe or inappropriate requests
  • Each question is tied to a specific claim and isolated from other claim information through the RISX-FACS security model
  • TLS encryption for data in transit 
  • Encryption at rest 
  • Role-based access control 
  • Private Azure tenancy 
  • Azure Key Vault for secret management 
  • Azure API Management as the API gateway layer 
Multimodal Input for Natural Language Questions Clients can ask specific claim questions by typing or speaking, without building reports or learning query syntax.
  • Users ask questions in plain language and receive answers without needing filters, report builders, or special training
  • Voice input makes the experience easier on mobile devices and more accessible for users who prefer or need hands-free interaction
  • Natural-language-to-SQL layer for structured data 
  • Hybrid AI search to retrieve relevant claim notes and supporting information 
  • Secure voice-to-text using Azure Speech to Text 

The next step in our generative AI journey

GB AI Assistant builds on our continued investment in generative AI and our focus on developing practical technology that enhances the everyday claims and risk management experience.

Alongside capabilities such as the Luminos Claim Summarizer, GB AI Assistant is part of our ongoing journey to use AI to simplify complex information, improve efficiency, and help our clients get more value from their claims data.

The goal isn't simply to introduce new technology. It's to make it easier for claim and risk professionals to access the right information, understand what it means, and act with confidence.

Ready to experience GB AI Assistant?

Discover how GB AI Assistant can help you spend less time searching through claims and more time making informed decisions.

Contact your Gallagher Bassett Client Service Manager to learn more or schedule a demo of GB AI Assistant.

Make Gallagher Bassett your dependable partner

When making the right decision at the right time is critical to minimize risk for your business, count on Gallagher Bassett's extensive experience and global network to deliver.

Connect with Us