What challenges did City of Montréal face before AI implementation?

Provide citizens faster access to public information and support Reduce call-center pressure during high-volume and seasonal spikes Connect citizens to relevant information across website content and internal systems

What was the impact of the AI solution?

High customer satisfaction in the first four months Over 85% of conversations handled with generative answers About 95% accuracy in internal testing Expected to help manage high call volumes and focus staff on emergency calls

Source

Microsoft

City of Montréal builds a Copilot Studio virtual agent for citizen self-service

Use case typeCustomer service automationUpdated Jun 13, 2026

The City of Montréal selected Microsoft Copilot Studio to build an AI assistant embedded in its public website. The assistant provides faster access to more than 40,000 pages of city information and integrates with two internal systems for waste collection and facility information. The city reports high customer satisfaction in the first four months and internal testing showed about 95% accuracy.

Organization: City of Montréal
Industry: Public Sector
Location: Canada
Published: August 2024

Reported outcomes

95%

accuracyQuality & accuracy

85%quantified impact

Strategic outcomes

Customer experience & trustImproved citizen self-service experienceNew product / capabilityLaunched website-embedded virtual agentSpeed & agilityProvided faster access to public informationScale & capacityReduced pressure on call center

Catalog median for quality & accuracy deployments: +90% across 282 reported metrics. Compare benchmarks →

Primary read

Use case focus

Showing 3 of 3

1Conversational AI for Citizen Services
2Virtual Agent
3Self-Service

Provide citizens faster access to public information and support
Reduce call-center pressure during high-volume and seasonal spikes
Connect citizens to relevant information across website content and internal systems

Microsoft Copilot Studio virtual agent embedded in the public website
Generative answers combined with pre-built chatbot responses
Integration with internal systems through existing APIs
Use of domain-specific entities such as postal codes and borough lists
Power BI dashboard and Power Platform deployment/pipelines

Technologies

Copilot Studio Power BI Power Platform

High customer satisfaction in the first four months
Over 85% of conversations handled with generative answers
About 95% accuracy in internal testing
Expected to help manage high call volumes and focus staff on emergency calls

Architecture

The City of Montréal deployed a Copilot Studio virtual agent on its public website. The assistant combines generative responses with pre-built chatbot responses, uses existing APIs to connect to backend systems, and leverages custom entities such as postal codes and borough lists. Analytics are tracked in Power BI, and the city plans to use Power Platform pipelines for more scalable deployment.

Sources & evidence1

Groundedness: 5/5Type: Customer StoryPublished: Aug 9, 2024Publisher: MicrosoftEvidence: PrimaryConfidence: High

Primary source

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Customer evidence