Junior AI/ML Engineer

Your Mission at Fourthline
Join our mission to prevent financial crime by developing best-in-class AI. You'll work alongside an exceptional and talented team, dedicated to making our onboarding AI the most reliable in the industry through cutting-edge technology. We're seeking a passionate and high-performing individual who is eager to contribute to our efforts. In return, you'll receive strong support from our collaborative AI team and mentorship from one of our senior AI engineers.

As a Junior Machine Learning Engineer, in the first six months your mission will be:

• Perform penetration testing on our mobile SDKs and AI systems by creating and injecting deepfake videos into our identity verification flow.
• Conduct research on the latest advancements in deepfake creation, mapping out current methods and assessing their levels of sophistication.
• Evaluate the risks associated with deepfake attacks.

After completion, you will shift your focus to these tasks:

• Support our AI team by building new models to classify user authenticity using RGB, depth, accelerometer, gyroscopic data, and more.
• Create prototypes and products using Python, TensorFlow, OpenCV, and more. You can add new tools if they suit our goals!
• Validate your experiments through rigorous testing on our extensive in-house dataset.
• Write code that adheres to our internal standards for style, maintainability, and best practices for a large-scale machine-learning environment. Maintain and advocate for these standards through code review.
• Contribute to thorough documentation—crucial for ensuring success and minimizing questions.
• Employ both technical and soft skills to address challenges, recognizing that solutions often require a blend of technology-based and people-based strategies. We work closely with our AI DevOps team, Fraud team, and Mobile teams to ensure your models add value across the board.
• Engage in continuous learning. We conduct monthly knowledge-sharing sessions to stay up-to-date on the latest research, new techand fundamental concepts in computer science.

Hard requirements (knock out criteria)

Required Qualifications:

• A completed MSc or Ph.D. in Computer Science, Computer Vision, Machine Learning, Math or a related field.
• Strong foundations in Python programming.

• 1+ years of software development experience in one or more general-purpose programming languages.
• Basic familiarity with Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs).
• Basic understanding of vector spaces and partial derivatives.
• Basic understanding of evaluation metrics such as accuracy, recall, precision and ROC AUC.
• Familiarity with machine learning frameworks such as Keras, TensorFlow, PyTorch, or JAX.
• You are either an EU citizen or in possession of a valid Dutch work visa (preferably already based in the Netherlands)
• You are willing and able to work in the office 3 days per week, due to the sensitive nature of the data you will be processing, which can only be accessed in our secure office environment.

Other things we would like to see

• In-depth understanding of deep learning algorithms and workflows, especially in handling large-scale visual data.
• Up-to-date knowledge of algorithms and models used in computer vision and machine learning.
• Experience in creating REST APIs and SDKs.
• Strong analytical and problem-solving skills, with the ability to think creatively and anticipate challenges.

Career Growth and Development 

At Fourthline, we offer opportunities for career progression and learning, allowing you to work alongside and learn from experienced engineers. As you grow within our team, there are possibilities to grow on our career ladder (we have Staff Engineer level and beyond), with clear pathways to further your technical, leadership and communication competencies. 

Our Stack 

Our technology stack includes Python as our primary programming language, with TensorFlow used for model development and OpenCV for non-model vision processing. We utilize Docker for containerizing our services and FastAPI for exposing service APIs. For managing Python dependencies, we rely on Conda, while AWS serves as our cloud hosting provider. To monitor our services, we use DataDog.