CANAL+ Group: Generative AI chatbot and AI Factory using Amazon Bedrock (training + POCs)
CANAL+ Group modernized its AWS environment to accelerate generative AI adoption for the myCANAL platform, including speech-to-text, natural-language search, and chatbot enhancements. Working with AWS Training and Certification and AWS experts, the company built cloud skills, ran a hackathon for rapid proof-of-concepts, and applied guardrails for responsible generative AI. CANAL+ also created an internal AI Factory for the Gender Equality Monitoring framework, using Amazon S3 and ML workflows to surface insights on gender representation.
- Organization
- CANAL+ Group
- Industry
- Tech & Comms
- Location
- France
- Published
- May 2026
Reported outcomes
−70%
timeTime & speed
Strategic outcomes
Catalog median for time & speed deployments: −60% across 727 reported metrics. Compare benchmarks →
Primary read
Use case focus
Showing 3 of 5
- 1Generative AI Chatbot
- 2Customer Service Automation
- 3Knowledge Management
- Modernize the AWS environment and support new digital media services at scale.
- Accelerate generative AI adoption for myCANAL while addressing hallucination and data privacy concerns.
- Build internal cloud skills to improve delivery speed and cost efficiency.
- Used AWS Training and Certification and AWS expert support to upskill staff and modernize the AWS environment.
- Ran a hackathon to rapidly develop POCs for speech-to-text, natural-language search, and chatbot enhancements.
- Built a generative AI chatbot using Amazon Bedrock and implemented guardrails for responsible AI.
- Established an AI Factory / Gender Equality Monitoring framework using Amazon S3 and ML workflows.
- Projects delivered 60-70% faster.
- IT costs reduced by 30%, totaling about $1.09M (€1M).
- About three months saved in getting the generative AI chatbot platform ready.
Implementation partners2
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?