Staff ML & AI Scientist

The Staff ML Scientist will collaborate with a team to conduct world-class applied AI research on financial payments data, driving innovation in alignment with Visa's strategic vision by incubating new data- and AI-powered products and enhancing existing applications with machine learning and AI. This role represents an exciting opportunity to make key contributions to Visa's strategic vision as a world-leading data-driven company. The successful candidate must have strong academic track record and demonstrate excellent statistical, machine learning and software engineering skills. You will be a self-starter comfortable with ambiguity, with strong attention to detail, and excellent collaboration skills. Essential Functions - Develop and apply cutting-edge algorithms and models, ranging from classical machine learning to deep learning techniques, including advanced neural network architectures such as Transformers, Graph Neural Networks (GNNs), and other emerging paradigms. - Pioneer and apply novel data science, deep learning, and AI methodologies to address unique business challenges and drive innovation. - Stay up-to-date with the latest research in machine learning, deep learning, and neural network architectures, integrating relevant advancements into business solutions. - Build, experiment with, and implement statistical, machine learning, and deep learning algorithms - including custom techniques as well as industry-standard tools. - Devise and apply advanced methods for explainability and interpretability of deep learning models, including mechanistic interpretability and model transparency techniques. - Develop and implement adaptive learning systems, as well as methods for model validation, A/B testing, and robust performance evaluation. - Collaborate with data engineers, software developers, product teams, and business stakeholders to translate business requirements into impactful machine learning solutions. - Communicate complex technical concepts, findings, and recommendations clearly to both technical and non-technical audiences. - Work with both structured and unstructured data, experimenting with in-house and third-party datasets to evaluate their relevance and value for business objectives. - Automate all stages of the predictive pipeline to streamline development and minimize manual intervention in both development and production environments. This is a hybrid position. Expectation of days in office will be confirmed by your Hiring Manager. Qualifikationen Basic Qualifications -MS or PhD in a quantitative discipline such as Statistics, Data Science, Mathematics, Physics, Operations Research, Engineering, or a related field, with demonstrated strength in machine learning, deep learning, or equivalent practical experience.
 Preferred Qualifications -7+ years of directly related experience applying data science and machine learning to solve business problems, with proficient Python coding skills and deep expertise in statistical analysis.

 -Exceptional problem-solving abilities, with experience designing and implementing complex data science solutions. -Hands-on experience developing and deploying deep learning models using PyTorch, including model architecture design and optimization. -Strong background in deep learning, including architectures such as Transformers. Experience with Large Language Models (LLMs), natural language processing (NLP), and advanced expertise in time-series modeling techniques. -Proficiency with big data tools and frameworks (e.g., Spark, Hadoop), and practical experience implementing MLOps practices such as model versioning, automated deployment, and production monitoring. -Strong understanding of model interpretability techniques, with the ability to analyze, articulate, and justify the decision-making processes of machine learning and deep learning models. -Experience working with financial data and building machine learning solutions for financial services, trading, risk, or related applications is desired. -Publications in recognized machine learning, data mining, or artificial intelligence journals and conferences are a strong plus.