Waste Management

AI answers waste-collection questions and returns location-specific schedules from back-end systems or public content.

Business impact

• Conversation automation rate — In Montréal, over 85 percent of conversations are handled with generative answers connected to the public website.
• Indexed documents — Montréal tripled the number of indexed documents within three months after improving Bing indexation.

Data requirements

• Public website content (text) — The agent answers from city website pages.
• Postal code (tabular) — Custom entities help identify the correct location.
• Street number (tabular) — Location details are used to retrieve the relevant schedule.
• Back-end system data (tabular) — Internal systems provide customized waste-collection and location-specific information.

AI methods and techniques

• Conversational question answering — The agent takes natural-language questions about waste collection and responds directly.
• Retrieval from indexed web content — It searches public website content through Bing indexation to ground answers.
• API-connected lookup — It calls back-end systems to retrieve customized waste-collection schedules and other location-specific information.
• Generative summarization — A language model summarizes the retrieved response into a user-facing answer.

AI models and model families

Microsoft Copilot Studio

Industries

Real-world evidence

2 documented case studies on record.

Companies using this: Waste Management (WM), City of Montréal.