Publicada el 15 junio
Misión del puesto
Overview
As part of Santander’s general AI Transformation team, you will lead strategic AI initiatives, connecting business, AI, and technology to drive measurable value. The Data & AI Science (DAISci) team develops, tests, and validates emerging AI methods into robust, compliant production systems across risk, customer experience, operations, and beyond.
This Scientist Responsible AI role focuses on ensuring AI is used responsibly, safely, and aligned with corporate standards, bridging scientific research and applied execution.
Main Responsibilities
You will contribute to the development and operationalization of Santander’s AI principles, working across DAISci areas to build safe, trusted, and responsible AI throughout the product lifecycle—from design to deployment and monitoring. The role combines scientific research with applied execution, advancing responsible AI methodologies, interpretability techniques, robust frameworks, and alignment strategies, and translating them into scalable, fair, and reliable systems. You will collaborate with engineering, validation, governance, and product teams; contribute to Santander’s research and publication agenda; explore emerging approaches in trustworthy AI; and mentor peers.
Qualifications
Education (Required): Advanced degree (PhD or MSc) in Computer Science, Artificial Intelligence, Mathematics, Statistics, or related field with specialization in Responsible AI, trustworthy machine learning, AI safety, fairness, interpretability, or AI governance.
Professional Experience (Required): Proven track record of advancing AI/ML through high-impact research publications, patents, open-source contributions, or transformative product innovations—particularly in areas related to model robustness, fairness, explainability, privacy, or safety.
Experience translating research into production (Required): Demonstrated ability to operationalize concepts such as bias detection and mitigation, interpretability techniques (e.g., SHAP, counterfactuals, causal methods), robustness testing, model monitoring, uncertainty estimation, or safety guardrails into scalable and reliable systems.
AI Red Teaming (Required): Experience in AI red teaming, including vulnerability analysis of conversational AI systems (LLMs), prompt injection, jailbreak strategies, and security testing of generative AI applications.
Benchmarking & Evaluation (Required): Experience conducting benchmarking and evaluation of AI/ML models, with particular focus on bias, fairness, and robustness assessments in both traditional ML systems and LLMs.
Hard Skills (Required):
- Strong programming skills in Python and familiarity with frameworks such as PyTorch, TensorFlow, DGL, or PyG.
- Deep understanding of responsible AI principles, including fairness metrics, model risk management, explainability techniques, adversarial robustness, privacy-preserving methods, and regulatory alignment (e.g., AI governance frameworks).
- Strong connections with academia and the wider AI research community, a passion for shaping the frontier of responsible AI, and ability to shape research agenda.
Soft Skills (Preferred): Ability to collaborate effectively with both technical and non-technical stakeholders.
Benefits
- Competitive salary with performance-based bonuses.
- Preferential banking terms, special interest rates on loans, life insurance.
- Health and wellness program (BeHealthy).
- Childcare support and family-friendly programmes.
- Legal, emotional, and administrative advisory services (Santander Contigo).
- Gym/WellHub membership, medical centers, meal subsidy, parking, shuttle service, exclusive employee discounts.
- Hybrid working model with flexible hours.
- Access to global learning platforms and courses.
Equal Opportunity and Accessibility
Santander is proud of being an organization where there are equal opportunities regardless of age, gender, disability, civil status, race, religion or sexual orientation. We are committed to providing an inclusive and accessible application process for all candidates.
#J-18808-Ljbffr
