Normalized claim
Quantified impact: 90%
Generated 11 million conversations and 90% positive feedback.
Sanofi builds Concierge, an agentic generative AI assistant on Amazon Bedrock
Use case typeAI agentsUpdated Jun 16, 2026
Sanofi built Concierge in-house on AWS as a unified, agentic AI companion for its global workforce, using Amazon Bedrock for model flexibility and planning to add Amazon Bedrock AgentCore for orchestration. The system gives employees a single conversational interface for finding information, generating content, completing tasks, and discovering specialized agents across the company.
Reported outcomes
90%
quantified impactOther quantified impact
−90.5%time−78%time−77%time−92%time
Strategic outcomes
Speed & agilityTransformed internal work into faster digital workflowsCustomer experience & trustImproved employee trust in AI-supported workScale & capacityCreated a scalable enterprise AI gateway for the workforceInnovation & cultureEncouraged internal app and agent building
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
Time: 90.5% decrease
Surveyed employees reported 90.5% time savings on routine tasks, 78% better decisions, 77% improved work-work balance, and one use case cut verification time by 92%.
Normalized claim
Time: 78% decrease
Surveyed employees reported 90.5% time savings on routine tasks, 78% better decisions, 77% improved work-work balance, and one use case cut verification time by 92%.
Normalized claim
Time: 77% decrease
Surveyed employees reported 90.5% time savings on routine tasks, 78% better decisions, 77% improved work-work balance, and one use case cut verification time by 92%.
Normalized claim
Time: 92% decrease
Surveyed employees reported 90.5% time savings on routine tasks, 78% better decisions, 77% improved work-work balance, and one use case cut verification time by 92%.
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Sanofi
- Provider
- AWS
- Maturity
- Production
- Linked source
- AWS
Deployed across multiple regions to reduce latency
Customer identity supportedSource describes one deploymentMaturity supported
Primary read
Use case focus
Showing 3 of 3
- 1AI agents
- 2Employee productivity
- 3Workflow automation
- Scale an internal AI companion across a global pharmaceutical workforce while preserving flexibility and strong data sensitivity controls.
- Move beyond basic summarization as employee requests became more complex and the tool became a primary digital gateway.
- Built Concierge in-house on AWS using Amazon Bedrock and a model-agnostic architecture.
- Deployed across multiple regions to reduce latency.
- Added an agent hub/orchestration layer connecting the companion to specialized agents for data retrieval, system write-back, and workflow automation.
Technologies
- Reached 50,000 weekly users and 72,000 monthly users.
- Generated 11 million conversations and 90% positive feedback.
- Surveyed employees reported 90.5% time savings on routine tasks, 78% better decisions, 77% improved work-work balance, and one use case cut verification time by 92%.
Architecture
Sanofi built Concierge in-house on AWS as a model-agnostic AI companion using Amazon Bedrock. The architecture was deployed across multiple regions for lower latency and expanded into an agent hub/orchestration layer that connects the assistant to specialized agents for data retrieval, system write-back, and workflow automation. Sanofi plans to incorporate Amazon Bedrock AgentCore to build, connect, and optimize agents securely at scale.
Sources & evidence1
Evidence: Medium65/100Evidence strength
- Customer explicitly identified
- Deployment status explicitly supported
- Primary source available
- Quantified outcome available
- Technical implementation details available
Type: Customer StoryPublished: Jun 16, 2026Publisher: AWSEvidence: 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
No published comments yet.