Grupo Cajamar and IBM have announced a new agreement to build a data and artificial intelligence (AI) platform aimed at automating processes, strengthening control over information, and making it easier to use AI models and agents across the bank’s operations. The infrastructure will combine systems installed on Cajamar’s own premises with cloud and open-source technologies, with the goal of serving an institution with more than 4 million customers.
The Cajamar-IBM platform in 30 seconds
- Cajamar and IBM will build a data and AI platform for banking processes.
- The architecture will combine IBM Fusion HCI, the cloud, Red Hat OpenShift, and IBM watsonx.data.
- IBM watsonx will enable AI models and agents across business processes.
- Cajamar is already piloting conversational agents to speed up fraud management.
- A joint team built an internal application in 120 days using IBM Bob, doubling the planned functionality.
The project aims to centralize Grupo Cajamar’s data in a governed environment with an open architecture. According to information provided by both companies, this setup will allow information to be processed more efficiently and securely, and will make it easier to gradually roll out AI capabilities across different processes at the institution.
The technology infrastructure will rest on IBM Fusion HCI systems installed on Cajamar’s premises, combined with cloud capabilities. On top of that, the bank will use Red Hat OpenShift, an open-source platform that, according to the companies, adds flexibility and keeps the architecture from being locked into a single vendor.
The data layer will run on IBM watsonx.data. Cajamar intends to use this environment to unify its data and strengthen governance, while keeping an architecture that can adapt to business needs.
IBM watsonx brings AI into banking processes
The new platform will serve as the foundation for developing and deploying AI models and agents through IBM watsonx. The goal is to embed these capabilities directly into business processes and automate repetitive tasks, freeing up staff to spend more time on work that adds value for customers.
Cajamar is already running pilot projects in this area, including the development of conversational agents aimed at improving the speed of fraud management. The information published by the bank doesn’t yet detail the results of these pilots or specify when they’ll move into production.
The agreement also includes IBM watsonx.governance for managing and overseeing its AI systems. This technology is part of the institution’s strategy for establishing governance, security, and control mechanisms around its AI development and for helping meet applicable regulatory requirements.
This point carries particular weight at a financial institution, where the use of AI models touches data and processes that require specific controls. In the project presented by Cajamar and IBM, governance is built into the architecture from the earliest stages, rather than being treated only as a layer added after development.
Mercedes Aranda, general manager of People and Technology at BCC-Grupo Cajamar, said the architecture will make it possible to work on internal processes and free up time to develop products and services tailored to different customer profiles.
For IBM’s part, Horacio Morell, president of IBM Spain, Portugal, Greece, and Israel, presented the project as an example of technology transformation in the financial sector. That assessment comes from the company and isn’t an independent evaluation of the project.
IBM Bob speeds up software development at Cajamar
The collaboration also extends to application development. Teams from Grupo Cajamar and IBM have started using IBM Bob, an AI-agent-based development tool that supports different stages of the software lifecycle.
According to information provided by the companies, IBM Bob is used for tasks including planning, coding, testing, deployment, maintenance, and application modernization. The tool also includes mechanisms related to governance, security, and cost control.
One concrete result Cajamar and IBM attribute to this collaboration is the development of a new internal application in 120 days. The joint team reportedly doubled the functionality originally planned for the project.
The 120-day figure applies to the specific case described by the companies and isn’t by itself a general benchmark for how long it takes to build applications with IBM Bob. The companies also don’t detail the size of the team, the volume of code produced, or a comparison with how long the same project would have taken without this technology.
The initiative reflects a strategy in which AI isn’t limited to the services the bank offers customers. It’s also being applied to data management, internal processes, fraud detection and management, and software development itself.
For Cajamar, the stated goal is to have a common platform from which to expand these uses while maintaining controls over data and AI systems. The open architecture built on IBM and Red Hat technologies is also meant to preserve technological flexibility as AI models and tools continue to evolve.
The project is therefore at a stage where already-deployed infrastructure, pilot projects, and tools teams are using in specific developments all coexist. Cajamar’s announcement doesn’t detail which other banking processes will be automated, nor does it lay out a full timeline for extending AI agents across the entire organization.
The institution has more than 4 million customers and positions this platform as a technology foundation for integrating artificial intelligence across different areas of its business. How the project evolves will determine how far today’s pilots and developments turn into everyday services and processes.
FAQ
What are Cajamar and IBM building?
The two companies are building a data and artificial intelligence platform to integrate AI models and agents into Grupo Cajamar’s business processes.
What technologies are part of the platform?
The architecture combines IBM Fusion HCI installed at Cajamar’s facilities with cloud technologies, Red Hat OpenShift, IBM watsonx.data, and AI governance tools such as IBM watsonx.governance.
What will Cajamar use AI agents for?
The institution is exploring their use across different processes. One of the pilot projects mentioned is the development of conversational agents to improve fraud management.
What results has Cajamar gotten with IBM Bob?
According to Cajamar and IBM, a joint team built an internal application in 120 days and doubled the functionality originally planned for the project.
