Your passion: data, AI and ML. Our goal: transforming Fin into Tech. Your future team, staffed with 15 experts, pursues this goal every day – by working on interdisciplinary projects and applying AI know-how to a range of possible application domains. You value international exchange at the highest level, think outside the box and enjoy sharing your knowledge? So do we! Join ING and stay one step ahead.
Your Tasks:
To solve specific problems, you leverage modern machine learning techniques as well as the vast amounts of data available within the organization. Primarily, you design and/or select the best modeling solution for a given business case, control the complexity of the model and data dependencies. Plus, you choose, build, tune and evaluate machine learning models and help integrate them. When faced with a new problem, you know which methods to try first and what trade-offs are involved. By anticipating them, you find the right balance between model stability, predictive performance, fairness and explainability. You also consider privacy and ethical topics surrounding the application of AI to personal data. Add to our team and quench your thirst for learning about new techniques and tools and pushing the envelope!
Your Profile:
- Master’s degree or PhD in computer science and several years of experience in building machine learning models
- Solid experience with Python and its data science libraries (scikit-learn, pandas, NumPy, TensorFlow, PyTorch)
- Versed in Natural Language Processing and its techniques: traditional (tf-idf, topic modeling) & recent (Word2vec, BERT)
- Know-how in Big Data platforms & tools (e. g. Apache Spark)
- Experienced in software development and best practices in a development team (testing, CI/CD, code reviews, etc.)
- Entrepreneurial mindset, passion to explore, creativity
- Ready to travel – up to max. 1⁄4 of your working time
- Fluent in English and willing to learn German
Apply now at Germany’s third largest retail bank:
Link: https://www.ing.jobs/Deutschland/Stellenangebote.htm