Success case · CaixaBank

CaixaBank: a 4-year Data & AI School, built for 6,972 employees

Ironhack won CaixaBank's corporate AI RFP as sole training partner, and is now running an in-house tech school across three specialized programs: AI, Business Analytics and Digital Transformation.

6,972 employees4-year program

CASE STUDY · SOURCE: IRONHACK ACCOUNT MATERIALS

Colleagues reviewing a learning path togetherCAIXABANK TECH SCHOOL · ILLUSTRATIVE PHOTO
The challenge

Uneven AI and data skills, and a fragmented supplier map

Knowledge of AI and data analytics varied widely from one area to another, and a patchwork of training vendors meant isolated initiatives with limited reach and no way to scale group-wide.

6,972

employees to be trained across the group.

69

nationalities, diversity as a catalyst for the change.

159.2h

of training per employee, per year.

4.8 / 4.9

out of 5: student and client satisfaction.

Ironhack × CaixaBank, Data & AI School
CaixaBank chose one partner over many, trading a patchwork of vendors for a single, coherent tech school with one curriculum, one platform and one place to measure impact.
The operating model

Six phases, run per course, on overlapping cycles

The same model AstraZeneca and Nestlé use for AI adoption, adapted here to run a full academy: análisis, diseño, desarrollo, formación, evaluación, iteración.

01

Analysis

Training needs, affected profiles, target KPIs.

02

Design

Content and method mapped to those KPIs.

03

Development

Materials and resources built for banking.

04

Delivery

Courses run, live and on-platform.

05

Evaluation

Results, satisfaction and attendance measured.

06

Iteration

Content refreshed twice a year.

Three schools

Every course built around a banking case, not a generic example

Credit scoring, ATM logs, card transactions, GDPR and Basel III compliance: the curriculum uses CaixaBank's own operating context throughout, not retail or generic use cases.

AI School

  • Fundamentos Generales de la IA, 42 hours.
  • Ética y leyes de la era digital, 30 hours.
  • Ingeniería de Prompt, beginner and advanced tracks, 26 and 24 hours.
  • Investigador de soluciones de IA, 42 hours, and NLP, 34 hours.

Business Analytics

  • Introducción al Business Analytics, 25 hours.
  • Toma de Decisiones y Ética del Dato, 27 hours.
  • Visualización de Datos y Storytelling, 25 hours.
  • Dirección Estratégica en Business Analytics, 30 hours.

Cross-school

  • Resolución de problemas complejos y pensamiento crítico, 30 hours, shared across tracks.
  • Delivered through the Ironhack Student Portal, with an AI Tutor giving real-time feedback.
  • Dedicated Slack channels between instructors and learners.
GovernanceThe relationship is personally overseen by Ironhack co-founder and CEO Gonzalo Manrique Sabatel, who runs course kick-offs, quarterly reviews and closing ceremonies.
Where this is headed

A productivity gain still being measured, not yet declared

Ironhack's cross-client benchmark points to a productivity improvement in the 30–35% range once a program like this is fully adopted. For CaixaBank specifically, that figure is being validated during the program's initial analysis phase, not assumed in advance.

The scale is real: 6,972 people, 69 nationalities, a 4-year commitment. The productivity number is the one thing being measured honestly rather than declared upfront, and that is deliberate.
Same model, different depth

Ready to build your own tech school?

Start with one course, or plan a multi-year academy.