Expert Data Scientist AI Factory - AI Driven Process Automation Excited to grow your career? BBVA is a global company with more than 160 years of history that operates in more than 25 countries where we serve more than 80 million customers. We are more than 121,000 professionals working in multidisciplinary teams with profiles as diverse as financiers, legal experts, data scientists, developers, engineers and designers.
Learn more about the area: We're a group of people with a passion for cutting edge technology, mutual learning and innovative problem solving. We strive to bring the opportunities that Artificial Intelligence offers to the financial industry.
At AI Factory, we are excited to welcome a Data Scientist to our team, dedicated to harnessing the power of LLMs in the financial industry. In this role, you will be at the forefront of managing and executing significant analytical projects within the DATA area. Your responsibilities will include coordinating with team members, aligning with our strategic objectives, and solving complex problems by working closely with cross-functional teams. You will play a key role in identifying business opportunities and applying your data science expertise to these challenges. Ensuring that all deliverables comply with Advanced Analytics discipline governance standards and best practices is an essential aspect of your role. Additionally, your experience will be vital in mentoring and guiding less experienced team members, helping them to grow in their roles and contributing to the overall success of our team.
Tasks Definition and development of the analytical cycle of solution evaluation:
Ability to assess and select appropriate LLM models for each business need. Execution of LLM test plans, including the definition, construction and obtaining of analytical metrics (performance/cognitive, technical, and business) capable of evaluating LLMs in both discriminative and generative modes. Refinement (fine-tuning) of LLM models on new domain data (discriminative and generative). Iterative construction of training / fine-tuning / validation datasets, evolving them based on the understanding metrics obtained and case analysis. Methodology approach for A/B testing and definition of requirements for experiments. Support for the construction and validation of KPIs. Experimental development of analytical solutions:
Construction of knowledge bases for the models, e.g. domain documents, unstructured information from the vertical, etc. Iterative construction of partial or complete analytical solutions, possibly integrating inferences from different models. Integration of the product and evaluation of its performance according to methodology. Execution of sprint-by-sprint analytical experimentation cycles. Specialist support to the cross initiative:
Technological prospection of advances and state of the art of current NLP (LLMs, foundational models, proprietary / open-source models, etc.). Experimentation support for specialized lines (e.g. fraud, marketing, UX, talent and culture, etc.). Dissemination, demonstration and transmission of technical concepts suitable for different corporate profiles. Training Bachelor's or Master's degree in Computer Science, Statistics, Mathematics, or a related field. Specific Knowledge Experience in the life cycle of ML models, including research, selection, adaptation, testing, training and monitoring / retraining. Advanced technical knowledge of LLMs: Discriminative / generative models, fine-tuning / prompt engineering, model optimisation, embedding elicitation and comparison, fine parameterisation, LangChain, etc. Knowledge and ability to define the metrics that would apply to each type and case. Data lifecycle management, including identification, prior exploration (EDA), transformation and adaptation to the model/problem, definition of training datasets, validation, testing, dataset tracing, etc. Knowledge of the risks inherent in NLP models, especially in their generative component (hallucination of facts, omissions, inherent biases, unpredictability, etc.). Experience in ML projects on cloud environments, preferably AWS. (Desirable): Experience in other cloud environments: Azure, Google Cloud Platform / Vertex AI. (Desirable): Experience in other on-premise environments (Datio / Datio Evo). Skills: Algorithms, Learning Management Systems (LMS) Administration, Machine Learning Algorithms, Machine Learning Model Management, Natural Language Programming (NLP)
#J-18808-Ljbffr