Skip to content

Published 16 Jun 2026· via AWS· Customer story

Sanofi builds Concierge, an agentic generative AI assistant on Amazon Bedrock

Sanofi · Pharma · France

Headline result

The source does not publish a single comparable figure. Reported results are listed below.

Summary

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.

The problem

  1. 01Scale an internal AI companion across a global pharmaceutical workforce while preserving flexibility and strong data sensitivity controls.
  2. 02Move beyond basic summarization as employee requests became more complex and the tool became a primary digital gateway.

What they built

  1. 01Built Concierge in-house on AWS using Amazon Bedrock and a model-agnostic architecture.
  2. 02Deployed across multiple regions to reduce latency.
  3. 03Added an agent hub/orchestration layer connecting the companion to specialized agents for data retrieval, system write-back, and workflow automation.

Reported results

  1. 01Reached 50,000 weekly users and 72,000 monthly users.
  2. 02Generated 11 million conversations and 90% positive feedback.
  3. 03Surveyed 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%.
  • Transformed internal work into faster digital workflows
  • Improved employee trust in AI-supported work
  • Created a scalable enterprise AI gateway for the workforce
  • Encouraged internal app and agent building

Source

AWS

Customer story · 16 Jun 2026

Read the original(opens in a new tab)

How this deployment compares

Computed from the library · data 10 Nov 2026

Sanofi has 3 deployment records in the library, from 2023 to 2026. It is one of 28 AI agents deployments, 1 of them in Pharma. Like 63% of records, its source gives no figure.

Sanofi in the library

3 deployment records from Sanofi, 2023–2026

  1. Oct 2023Sanofi personalizes cancer treatment and accelerates drug discovery using advanced AITherapeutics research
  2. Apr 2026Sanofi Digital Accelerator uses AWS SageMaker for predictive AI and deploys Bedrock for generative AI pilotsDrug discovery
  3. Jun 2026This record · AI agents
How Sanofi uses AI: company page

Same pattern, whole library

One of 28 AI agents deployments

  • Pharma accounts for 1 of them (4%), ranking #7 of 10 industries.
  • 16 records of this pattern came before it; 2 have been published since.
  • Only 8 of the 28 put a number on the result (29%).
Provider mix
  • Microsoft29%
  • AWS25%
  • GCP25%
  • IBM18%
Most-used technology
  • Azure OpenAI 5
  • Amazon Bedrock AgentCore 4
  • Microsoft Foundry 4
  • IBM watsonx Orchestrate 4
  • Gemini Enterprise Agent Platform 4
  • Gemini 4

Outlined: used in this record.

Shares describe this library, not the whole market.

AI agents analysis

Results against peers

No figure here. What peers report

Like 63% of records in the library, the source gives no figure. Peers that do quantify similar claims report:

  • Time & speed283 figures, all industries · middle half 30%–76%50%median reduction
  • Customer experience50 figures, all industries · middle half 20%–90%28%median increase
  • Adoption & scale8 figures, all industries · middle half 134%–1100%250%median increase

Peer figures are as reported by their sources; percentage claims only.

How we checked this record

Evidence tier: supported

A credible source with a named customer, but not the organization's own account.

Named organization
Sanofi
Who published the source
The organization itself
Deployment status
Production (confidence 95%)
Quantified result
The source reports no figure
Summary checked against source
5 / 5 agreement

Last re-checked 26 Jul 2026.

This record is evidence for

Insight · industryPharma dossierOne of 97 Pharma deployments. See which patterns dominate there and what they report.Insight · patternAI agents28 deployments share this pattern. See where it is used and how it is built.Trend · momentumRising pattern15 new deployments in the last six months, against 10 in the six before.

Same pattern, other organizations

All ai agents deployments
  1. 10 Jul 2026
    Tech & Comms

    Visma Success Story - AWS

    Visma · AI agents · AWS

    Visma needed a scalable way to apply generative AI and agent capabilities across more than 160 decentralized business units while maintaining governance and developer autonomy. Using Amazon Bedrock for production generative AI applications and agents at scale, Visma aligned AI work across product development, growth functions, and workforce enablement while keeping regulatory compliance and data security in place.

    AgenticScaled Production

    4,000 count
    employees upskilled
  2. 1 Jul 2026
    Other

    Blackroll builds an AI recovery coach with Gemini Enterprise Agent Platform on Google Cloud

    Blackroll · AI agents · GCP

    Blackroll needed to unify disconnected customer data and build an interactive AI coaching experience that updates guidance based on wearable signals. It consolidated e-commerce records and mobile app events into BigQuery, used Cloud SQL to ingest daily wearable data, and deployed a headless architecture on Cloud Run, Firestore, and Cloud Run functions. Using Gemini Enterprise Agent Platform and Agent Development Kit (ADK), Blackroll built autonomous coaching agents and used Agent Search to ground recommendations in verified scientific research and knowledge from its 5,000 physiotherapists. Antigravity was used to automate software development loops for faster implementation.

    AgenticProduction

    −20%
    software footprint reduction
  3. 29 Jun 2026
    Tech & Comms

    Kimi builds production AI agent infrastructure on Alibaba Cloud (ACK + ACS Agent Sandbox + Lindorm)

    Kimi · AI agents · Alibaba

    Kimi rolled out multiple agent-based products and built an end-to-end agent infrastructure with Alibaba Cloud to handle high-concurrency requests, fast sandbox startup, state preservation, secure isolation, and low-cost elastic scaling. The architecture combines Alibaba Cloud Container Service for Kubernetes (ACK), ACS Agent Sandbox, Lindorm multi-model database, MicroVM isolation, NetworkPolicy, and Fluid to support production AI agents and model training workloads.

    AgenticProduction

    −>50%
    sandbox startup time
  4. 29 Jun 2026
    Retail

    New Aim deploys 24/7 agentic operational agents with Gemini Enterprise Agent Platform

    New Aim · AI agents · GCP

    New Aim built an AI-native operating layer on Google Cloud to replace siloed spreadsheets and manual SKU/inventory calculations. It centralized transaction and logistics data in Cloud Storage and BigQuery, used BigQuery ML for demand and pricing models, and deployed 24/7 internal agents with Gemini Enterprise Agent Platform and Gemini Flash models. The system lets planners query complex metrics in plain English, keeps human-in-the-loop validation for qualitative adjustments, and runs applications on Cloud Run.

    AgenticProduction