Senior Expert Data Scientist GenAI (LLMs)

BBVA; Madrid; On-site; EUR 90,000 - 130,000; Full time

Job summary

BBVA AI Factory en Madrid busca un Científico de Datos para liderar soluciones de IA Generativa y LLM, colaborando con equipos multidisciplinares para transformar problemas de negocio en productos de datos. La posición implica actuar como referencia técnica en Generative AI dentro del dominio y guiar el desarrollo de modelos en un entorno de banca avanzada.

Se requiere experiencia en ML/LLM, liderazgo y habilidad para comunicar hallazgos complejos a stakeholders.

Qualifications

• Bachillerato/Máster en Informatica, Estadística o Matemáticas (o campo relacionado).
• 8+ años de experiencia como Científico de Datos o profesional de ML, con entrega de soluciones en producción.
• Conocimientos profundos de LLM, incluyendo engineering de prompts, embeddings y retrieval-augmented generation.
• Dominio de Python, Pandas, NumPy y scikit-learn; experiencia con PySpark, TensorFlow, PyTorch y Transformers.
• Gestión del ciclo de vida de ML/LLM desde definición de datos hasta monitorización.
• Liderazgo teórico y mentoría a practicantes menos experimentados.
• Conocimiento de riesgos de modelos y principios de AI ética, cumplimiento de normas G DPR, etc.
• Capacidad de comunicación para traducir conceptos técnicos a negocio.

Responsibilities

• Actuar como referencia técnica de Data Science y Generative AI en el dominio.
• Dirigir iniciativas analíticas de gran impacto desde planteamiento hasta despliegue en producción.
• Desarrollar y optimizar modelos predictivos y generativos usando ML y LLM.
• Diseñar componentes de soluciones LLM, incluida preparación de datos y estrategia de recuperación de información.
• Evaluar y optimizar soluciones LLM considerando exactitud, relevancia y coste.
• Gestionar el ciclo de vida de datos y modelos, desde definición hasta monitorización.
• Trabajar con PMs, data engineers y platform teams para entregar productos de datos listos para producción.
• Tomar decisiones de ciencia de datos y documentar trade-offs clave.
• Mentorizar a data scientists y realizar revisiones tecnicas.
• Comunicar hallazgos y limitaciones a stakeholders tecnicos y de negocio.
• Seguir normas de gobierno de Análisis Avanzadas y buenas prácticas de desarrollo.

Skills

LLM expertise

Python

Machine learning

Leadership

Communication

Ethical AI

Education

Bachelor's or Master's in CS/Stats/Math

Tools

Pandas

NumPy

scikit-learn

PySpark

TensorFlow

PyTorch

HuggingFace

LangChain

Job description

¿Te entusiasma hacer crecer tu carrera?

BBVA es una compañía global con más de 160 años de historia que opera en más de 25 países donde damos servicio a más de 80 millones de clientes. Somos más de 121.000 profesionales trabajando en equipos multidisciplinares con perfiles tan diversos como financieros, expertos legales, científicos de datos, desarrolladores, ingenieros y diseñadores.

Conoce más sobre el área:

At BBVA AI Factory…

BBVA AI Factory operates as a global hub within the Data area of BBVA, with development centers in Spain, Mexico, and Turkey.

Our mission is to build complete, end-to-end data products that solve BBVA's business needs by working closely with business units to transform strategic priorities into actionable, data-driven solutions. Some of our recent projects include:

• Mercury Library, an in-house AI framework now available to the entire data community, aimed at boosting collaboration and accelerating AI solution development. [https://www.bbvaaifactory.com/category/mercury/]
• A machine learning pipeline designed to enhance early debt recovery by predicting default risk and optimizing collection strategies [https://www.bbvaaifactory.com/a-machine-learning-pipeline-for-early-debt-recovery-executive-summary/]
• Applying daily life embeddings to drive deeper personalization in customer interactions and improve service recommendations. [https://www.bbvaaifactory.com/behind-daily-life-embeddings-in-action/]
• Utilizing conformal prediction to provide reliable uncertainty estimates and enhance the confidence in AI model predictions [https://www.bbvaaifactory.com/conformal-prediction-an-introduction-to-measuring-uncertainty/]
• Building algorithmic explainability frameworks to ensure transparency and foster trust in our AI systems. [https://www.bbvaaifactory.com/algorithmic-explainability-how-do-we-apply-it-in-bbva-ai-factory/ ]

At BBVA AI Factory, innovation isn’t just a goal‑it’s a continuous journey.

Sobre el puesto

At BBVA AI Factory, we're seeking an Data Scientist to join our dynamic team in Madrid, specializing in the application of Generative AI and Large Language Models (LLMs). Our team thrives on collaboration and innovation, tackling complex challenges in finance with state-of-the-art AI solutions. Whether it’s developing explainable algorithms, building Machine Learning pipelines to improve debt recovery, or leveraging LLMs for enhanced customer experiences, we focus on creating real-world impact. As the technical reference for Generative AI within your domain, you will lead the data science approach for end-to-end AI products that enhance key banking processes. Working with product managers, data engineers, ML engineers and other specialist teams, you will guide the development of robust LLM-based solutions and help develop a new AI-supported customer relationship model that benefits both customers and managers.

Key Job Responsibilities

• Technical Leadership: Act as the technical reference for Data Science and Generative AI within the domain, defining the analytical and modelling approach for complex business problems.
• Lead Analytical Initiatives: Lead high-impact analytical initiatives from problem framing to production deployment, defining objectives, success metrics, experimentation plans and model performance criteria.
• Advanced Model Development: Develop and optimize predictive and generative models using statistical, machine learning and LLM techniques.
• LLM Solution Design: Design the data science components of LLM solutions, including data preparation, knowledge-base design, retrieval strategies, prompt engineering, model selection and evaluation methodologies.
• LLM Evaluation and Optimization: Establish and apply rigorous methodologies to evaluate, fine-tune and improve LLM-based solutions, considering accuracy, relevance, reliability, latency and cost.
• Data and Model Lifecycle: Ensure sound practices across the data and model lifecycle, from dataset definition and exploratory analysis to testing, monitoring and model improvement.
• Cross-Functional Collaboration: Work closely with product managers, data engineers, ML engineers, platform teams and specialist control functions to deliver production-ready data products.
• Technical Decision-Making: Make and document key data science trade-offs, such as model selection, evaluation thresholds, data requirements and experimentation design.
• Technical Mentorship: Mentor data scientists and contribute through technical reviews, knowledge sharing and guidance on analytical and modelling best practices.
• Stakeholder Communication: Communicate complex analytical findings, model limitations and recommendations clearly to technical and business stakeholders.
• Best Practices Compliance: Ensure that data science deliverables follow Advanced Analytics governance standards and agreed development practices.

Your Qualifications

• Education: Bachelor's or Master's degree in Computer Science, Statistics, Mathematics, or a related field.
• Experience: 8+ years of experience as a Data Scientist or Machine Learning practitioner, including substantial experience delivering machine learning and LLM solutions to production. Demonstrated experience providing technical leadership in multidisciplinary teams is required.
• LLM Expertise: Deep technical knowledge of LLMs, including prompt engineering, fine-tuning, embedding models, retrieval-augmented generation, model evaluation and model optimization.
• Programming & ML Frameworks: Deep expertise in Python, statistical modelling and machine learning, with strong proficiency in Pandas, NumPy and scikit-learn, plus hands-on experience with PySpark, TensorFlow, PyTorch and HuggingFace Transformers. Experience with libraries such as LangChain or LangGraph is a plus.
• Data Lifecycle Management: Proven experience across the end-to-end lifecycle of ML and LLM models, from dataset definition and EDA to experimentation, testing, deployment, monitoring and iterative improvement.
• Technical Leadership: Proven ability to define data science approaches for complex problems, make sound modelling decisions, conduct technical reviews and mentor less experienced practitioners.
• Model Risk Awareness: Strong understanding of LLM limitations and model risks, including hallucination, bias, performance variability and limitations arising from data quality. Ability to identify, communicate and address these limitations in collaboration with relevant specialist teams.
• Applied Machine Learning: Deep knowledge of a broad set of machine learning techniques applied to solve complex business problems, including A/B testing for model performance.
• Communication Skills: Excellent ability to translate complex technical concepts into actionable business insights.
• Ethical AI and Responsible Data Science: Knowledge of ethical AI principles, data privacy laws (like GDPR, CCPA), and a commitment to responsible data science practices.

Nice to Have

• Previous experience in the financial industry.
• PhD in Computer Science, Statistics, Mathematics, or a related field.
• Cloud Computing: Experience in AWS, Google Cloud, or Azure for scalable data science solutions, including experience with Docker and Kubernetes.
• Experience in Causality: Knowledge and application of causal inference methods to identify and model cause-and-effect relationships in data.
• Experience with MLOps or LLMOps practices, including experimentation tracking, model and prompt versioning, automated evaluation and model monitoring.

Habilidades:

Empatía, Ética, Innovación, Orientación al cliente, Pensamiento proactivo