Snohomish County speeds training with Azure Speech and Foundry; improves accessibility with AI narration
Snohomish County deployed an easy-to-use learning platform built on Azure Speech to modernize training and internal communications while reducing manual effort and improving accessibility. The workflow lets HR and learning teams generate multilingual narration, update scripts without re-recording, and deliver subtitles and transcription through a no-code self-service experience in Azure AI Foundry. The county is also initiating Microsoft 365 Copilot Chat and exploring future agent-based scenarios in government cloud.
- Organization
- Snohomish County
- Industry
- Public Sector
- Location
- United States
- Published
- July 2026
Reported outcomes
−87.5%
training production timeTime & speed
Strategic outcomes
Catalog median for time & speed deployments: −60% across 724 reported metrics. Compare benchmarks →
Primary read
Use case focus
Showing 2 of 2
- 1Content generation
- 2Accessible learning
- HR training video production was manual and slow, taking about four hours or more for a 10-minute video.
- The county needed faster content creation and improved accessibility and multilingual learning.
- Implemented an AI-based learning/content creation workflow using Azure Speech in Azure AI Foundry to generate multilingual narration and speech enhancements.
- Enabled quick updates to narration text without re-recording.
- Incorporated subtitles, transcription, and responsible AI guardrails in a no-code/self-service experience.
- Reduced training production time by 87.5%.
- Freed about 840 hours annually for higher-value work.
Architecture
Snohomish County built a no-code, self-service learning platform on Azure AI Foundry using Azure Speech to generate multilingual narration, subtitles, and transcription. The county retains control and governance while HR and learning teams can update scripts without re-recording and reuse the platform across departments.
Sources & evidence1
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Was this useful?