Scroll down

IFCAM uses Adaptive Learning to boost its Scan’up

Camille Duchasseint |

Presentation of IFCAM 

IFCAM, the university of Crédit Agricole Group, contributes to the performance of a Group that cares about human development and personal accomplishment by giving men and women the means to acquire and develop their skills.

Defined as the “university for all”, IFCAM supports all employees in their skills development, no matter their seniority, job or role, by offering learning paths adapted to their needs that combine face-to-face, online and blended learning.

It is also the “university for all needs”: the learning catalogue covers all the skills required for jobs in the field of banking, insurance, and real estate.

IFCAM is a major player in the Group’s transformation, and it promotes the values and culture of Crédit Agricole. In addition to providing quality learning programs, IFCAM anticipates and supports future developments.


In the rest of this article, the IFCAM team answers the various questions asked.

1. What are the different phases of the project?

The first phase of the experiment was conducted by the “Data and AI in Learning” team, headed by Vanessa Dastugue, with the support of the technical teams and the occasional contribution of the learning teams.

For the second phase, the large-scale implementation of the Adaptive Learning Scan’up, a transversal project team was assembled to ensure its success:

  • The adjustment of the content was carried out by the Learning Design division led by Pascal Mollicone, with the support of the Group’s learning coordinators and business experts
  • The study and implementation of the technical prerequisites for a large-scale implementation were done by the teams of the Technology Division, under the coordination of Pierre Krauzman
  • Other teams were involved in in the marketing, communication and distribution of the innovative solution

What is a Scan’up?

The term “Scan’up” comes from the fusion of “Scan” which means “photograph” and “up” which means “progress”, “grow”.

The idea of Scan’up is to photograph the skills of employees at a given moment (“Scan”), and then to provide them with a learning path that is perfectly adapted to their needs to help them improve their skills (“up”).

The objective of this teaching method, designed and implemented by IFCAM, is to achieve ever more efficient learning according to each learner’s profile.

The way it works is very simple: each learner carries out a self-assessment of skills, in the form of a questionnaire, which allows to get recommendations of personalized learning materials. Once the learning path has been completed, a final assessment is carried out to measure the progress.


2. What is the objective of the project and why Domoscio’s Adaptive Learning solution was chosen?

We have decided to introduce Adaptive Learning to our L&D strategy in 2018. After the success of our Adaptive Learning MOOCs, we chose to enrich our Scan’up learning programs. 

Today, we have 25 Scan’ups covering the Group’s core business: advisory services in various fields (insurance, agriculture, etc.) and for different client types (individuals, professionals, etc.). 

The objective was to enhance these existing Scan’ups with Adaptive Learning to:

  • Evaluate each employee’s skills level more efficiently and accurately with an assessment composed of only the relevant and discriminating questions according to their profile.
  • Hyper-personalize the experience by targeting individual needs and providing a more granular learning response to improve the employee’s skills development.
  • Meet a strong expectation from the Group’s entities that need to have visibility on the skills of their employees and support to adapt their skills development plans.
  • Optimize the time needed to complete the Scan’ups.

The impacts of AI are more significant when it is implemented in a context in which skills development is time-consuming for employees, but they are still required to complete compulsory learning. The challenge is therefore to optimize the time dedicated to increasing skills by providing individualized and efficient solutions that target the needs of each employee.  

Our entities’ learning departments also have high expectations:

  • Offering a more detailed and realistic view of the mastery of skills in a given job
  • Helping the interpretation and analysis of the results at the end of the quizzes. This is an important challenge in terms of change management because we are transforming the way indicators are read: we no longer focus on the score since the number of questions chosen by the AI and their level of difficulty varies from one employee to another, but we analyze the proficiency on a skill in relation to the expected level (for the employee’s job and level of expertise).
  • Providing a seamless and user-friendly tool for employees, well integrated to existing learning processes. 

Besides, the pilot that we have carried out has highlighted the attractiveness of the solution and a better engagement from employees thanks to the AI-powered system.

3. What did you implement in 2021 and what were the results? 


2021 was the year we confirmed the relevance of Domoscio Hub for our learning environment, with the implementation of 4 Adaptive Learning Scan’ups with 7 pilot entities (respecting the requirements of GDPR).

These 4 Scan’ups, designed for employees specializing in Inheritance and Agriculture, allowed us to confirm the attractiveness of the solution:

  • Better targeted questions, with 9 to 12% of time saved per employee compared to a Scan’up without AI
  • Better employee engagement, with 70% of learning recommendations consulted, compared to 25% in a traditional AI-free system
  • An average increase of more than 10 points between the first and the final assessment.
Illustration Temps gagné pour acquérir les compétences cibles
9 %

of time saved per employee

schéma de la rétention de la connaissance
70 %

of learning recommendations consulted

We were able to learn lessons in terms of learning design (adapting quizzes based on questions analysis) and gather feedback on the tool to provide improvements in ergonomics and readability of indicators. 

It was also an opportunity to clarify the challenges in change management, by gathering feedback from learners and learning services. Indeed, although AI is attractive, it can also be quickly challenged if the results are not in line with what is expected: a substantial effort must be made to explain how AI works to avoid this pitfall.

Finally, even though the small number of testers did not allow us to take full advantage of the possibilities of AI, the pilot allowed us to collect data for the large-scale implementation.

4. What have been the main challenges?

To take full advantage of AI from the start, significant work has been done on existing data, in collaboration with Domoscio: extraction of data from IFCAM’s information system, analysis by Domoscio algorithms, interpretation of the AI analysis and formulation of recommendations in terms of learning design. This is a time-consuming process, but it provides great added value to the system: we want to make it sustainable after its large-scale implementation. By collecting the data as employees follow the Scan’ups and computing this data with AI at regular intervals, we can ensure a real process of continuous improvement. 

In this type of project, a significant amount of data is collected, processed, transmitted, stored: special attention has therefore been given to securing these actions, and we have had to adapt our processes to guarantee our compliance, particularly with the General Data Protection Regulation.

5. What are the project next steps?

Currently, we are in the phase of generalizing Adaptive Learning for all existing learning programs, which means 25 Scan’ups concerned by late 2022 – early 2023. The next objective is to systematize the creation of new Scan’ups with Adaptive Learning. 

This stage of generalization opens a significant project in terms of learning design:

  • Extracting historical data from Scan’ups to have Domoscio algorithms analyze it and define priorities for redesign (for instance, levels of difficulty not covered by existing questions)
  • Improve the structure of Scan’ups with a skills-based approach (and not just learning topics)
  • Granularize learning recommendations in line with this new structure
  • Redesign questions/answers to maximize the relevance of the question bank

At the same time, technical work is being done to streamline the L&D and HR environments with regards to the use of the solution 

Signed the IFCAM project team

Photo locaux IFCAM