Normalized claim
Queries resolved: 17,000 queries increase
In its first two months, it resolved more than 17,000 queries, with usage continuing to grow.
MicrosoftEvidence: Medium50/100
FM builds Azure OpenAI + Azure AI Search solution for engineering standards search
Use case typeSearch modernizationUpdated Jun 24, 2026
FM (property insurance) built a secure AI-powered search solution on Microsoft Azure to help engineers retrieve engineering standards faster and more accurately across tens of thousands of pages of complex PDFs, diagrams, flow charts, and tables. The solution uses Azure OpenAI in Foundry Models and Azure AI Search, applies structured chunking aligned to engineering logic, validates answers against ground-truth examples, and continuously evaluates performance through governance and feedback loops.
- Organization
- FM
- Industry
- Insurance
- Location
- United States
- Published
- June 2026
Reported outcomes
17,000 queries
queries resolvedOther quantified impact
6-10 minutestime saved per search
Strategic outcomes
Customer experience & trustStrengthened client experience through faster access to engineering guidanceRisk & compliancePreserved accuracy and governance in a high-risk engineering workflow
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
Time saved per search: 6-10 minutes decrease
With 6–10 minutes saved per search, multiplied across more than 1,500 engineers, FM estimates the solution frees up thousands of engineering hours each year
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- FM
- Provider
- Microsoft
- Maturity
- Unknown
- Linked source
- Microsoft Customer Stories
No explicit deployment-stage evidence found.
Customer identity supportedSource describes one deploymentMaturity evidence evaluated
Primary read
Use case focus
Showing 2 of 2
- 1Search modernization
- 2Content discovery assistant
- Engineers needed faster, accurate retrieval of engineering standards across tens of thousands of pages of technical content.
- An out-of-the-box chatbot would not meet FM's requirements for reliability, transparency, or accuracy in a high-risk engineering environment.
- FM worked with Microsoft and Spyglass MTG to develop a secure, production-grade AI search solution on Azure.
- The system structures and chunks content according to engineering logic, returns answers with relevant context, uses ground-truth validation, and is designed for multi-region availability and 15-second response times.
- The solution resolved more than 17,000 queries in its first two months.
- FM estimates 6-10 minutes saved per search across more than 1,500 engineers, freeing thousands of engineering hours annually.
- Faster access strengthens client experience.
Architecture
The solution runs entirely within Azure, using Azure OpenAI in Foundry Models and Azure AI Search. Content is carefully structured and chunked according to engineering logic rather than document format alone. Answers are validated against hundreds of known ground-truth examples and continuously evaluated through governance and feedback loops. The architecture includes centralized identity authentication, Azure data protection and governance capabilities, multi-region support, least-utilized-region routing, and automatic failover across Azure regions for resilience and high availability.
Implementation partners1
Sources & evidence1
Evidence: Medium50/100Evidence strength
- Customer explicitly identified
- Primary source available
- Quantified outcome available
- Technical implementation details available
Type: Customer StoryPublished: Jun 24, 2026Publisher: MicrosoftEvidence: PrimaryConfidence: High
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
AI patterns
Browse the catalog
Was this useful?
Community
Comments
Loading comments...