# Evidence projet 1 - Groupama Groupama is a large insurance company in French, which is also a faithful mainframe customer. Their aging estate was due for a refresh in 2022 : the z14 machines they used had reached end of marketing, and with it, critical DR capabilities, putting their DR at risk. In 2022, as the z Technical Specialist in charge of Groupama, I led the team working on the refresh. The project, worth 6.77M USD in hardware, has been divided in multiple phases over almost 3 years. During the capacity planning study, I evaluated multiple scenarios, active-active and active-passive, also looking at sub capacity models. At the time of the study, the client was, at the same time, implementing a datacenter relocation project, so the final architecture was not set at the beginning of the project. Therefore, customer asked us to study several options, to accomodate all decisions. After I presented it to the client, we agreed on the appropriate target models. Yet, the final datacenter architecture still undecided, the client insisted I keep the solution flexible enough to cater with all potential models. As such, I configured and explained the Flexible Capacity for Cyber Resiliency was THE exact feature for their needs, giving them the means to change for active active to active passive models without IBM involvement. Over the years, IBM's relationship with Groupama has been complicated. Through regular communications, transparency, rigorous work and quality documentation, I managed to turn the relationship around during the z16 refresh project. Below message sent by the client at the end of the technical study is a testimony : "Bjr, Je tiens à vous remercier pour la grande qualité du travail fourni par l'équipe IBM sur l'étude du sizing Z16. Nous sommes maintenant dans la dernière ligne droite et tous les éléments remontés permettent je pense d'avoir une trajectoire sécurisée. Merci de passer le message à toute l'équipe. O. Pastori, Responsable Ingénierie des Infrastructures Techniques Mainframe, Groupama" Once the contract was signed late June 2023, I kept working closely with the client : - shortly after the signature, he called me, upset. Despite regular communication and reviews, a misconfiguration happened. I solved the issue teaming up with the Client Care and Product Engineering teams to replace the faulty devices prior to the power on of the machines. - I also accompanied the client in the deployment and use of Flexible Capacity. I recorded a video illustrating the setup process, and I also ensured tight communication between our SSR onsite, PE and client team to ensure smooth installation, as it was a first in France (and in Europe!). As planned, the client successfully used the functionality during its datacenter move project in the summer of 2025. One of the lessons I learned is that regular, factual communication goes a long way. It allows me to identify and solve difficulties quickly and early, before they turn into critical situations. # Evidence project 2 - CA-GIP Credit Agricole is the 2nd largest French bank by asset sizes, the 09th worlwide. The bank operates a very large mainframe estate to deliver services to their customers, operated by CA-GIP (Credit Agricole Group Infrastructure Platform). At the end of 2Q2023, we sold them a LinuxONE machine, part of a 6.7M USD deal. This was the culmination of several months of work where I developed the platform's value proposition together with a total cost of ownership analysis, to position LinuxONE as an ideal platform to modernize and innovate close to the mainframe. The lines of business at the bank were interested to evaluate the benefits of using LinuxONE for their Openshift infrastructure to host critical, customer- and staff- facing applications. Since the expectations were very high, we started right from the beginning to define a project to get the project going, aligning all relevant stakeholders behind a shared goal. I laid out a high-level plan to move the project forward. Client nominated a project manager, with whom I interlock on a weekly basis to follow progress, identify and overcome risks as they arise. To ensure the LinuxONE test at Credit Agricole is a success, I organized, and still lead, different work streams : - architecture : to define the required deployment model for the PoC, as well as the resiliency scenarios. I led this stream with the help of Client Engineering, engaging all relevant stakeholders regularly to discuss and agree on options. - deployment : once the architecture was defined. I worked closely through planned and unplanned meetings, with the distributed teams in charge of the deployment and management of the platform. - ecosystem : engaging the IBM z WW Ecosystems team to ensure I can give clear visibility to the client about the availability of their required software on the platform. - network : I involved the lab to address some limits of Openshift with regards to leveraging the RoCE adapters on the platform. In April 2025, I invited one of the architects to an IBM TechXchange event in Amsterdam about Containers on IBM Z. I took this opportunity to introduce him to many of the teams involved in Openshift architecture, development and support. I used this event as the perfect occasion to trigger discussions with SMEs to comfort some of the client's choices, exclude others and put our work into longer term perspective. I developed close intimacy with the client, who is heavily involved in the project. We now regularly share information, even outside the scheduled meetings. This project is ongoing. Thanks to our close collaboration, early September, the client has successfully tested their first application on Openshift on LinuxONE. Further testing is in progress, and the second use case around AI will be deployed shortly. However, what I took away from this project is how critical it is to rely on the broader IBM to succeed in such complex project with a lot of moving parts.