Air France-KLM
Generative AI factory on Google Cloud with a private assistant and maintenance RAG
Air France-KLM built a generative AI factory with Accenture and Google Cloud and reported development cycles over 35% faster
Problem
Air France-KLM wanted a common way to build, test and run generative AI across the group, from ground operations and engineering to customer service.
Approach
Accenture helped the group build a generative AI factory on Google Cloud. Teams use it to develop, test and manage a catalogue of pre-evaluated generative, agentic and machine-learning models. A private generative AI assistant is already in use. Retrieval-augmented generation supports diagnosis and repair of aircraft damage. The group also ran GenAI Days, where employees learned the tools and proposed ideas.
Results
Takeaway
The partners said results had appeared in ground operations, engineering and maintenance, and customer service. They gave no figures for those results and no method for the development-speed claim.
Categories
- Size
- Large enterprise
- Industry
- Travel and aviation
- Business function
- IT and R&DProduction and quality
- Innovation type
- Process innovation
- AI technology
- Generative AIAI agents
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
Accenture's release describes a generative AI factory built on Google Cloud for Air France-KLM, with a private assistant and RAG for aircraft damage diagnosis and repair. It says development cycles sped up by over 35% and reports results in several areas without metrics. It includes quotes from Accenture, Google Cloud France and Air France-KLM.
Full details are in the original sources.