When sourcing an ML engineer, a deep knowledge of machine learning and deep learning algorithms is crucial. The candidate should demonstrate experience working with different types of learning, including supervised, unsupervised and reinforcement learning, and be familiar with popular models such as neural networks, decision trees, clustering and regressions.

It is important that the candidate has experience with popular ML libraries and tools such as TensorFlow, PyTorch, Keras, Scikit-learn. Knowledge of these tools is essential to effectively design, train and implement ML models.

Solid programming skills, especially in Python or R languages, are essential for an ML engineer. The candidate should be proficient in writing clean, modular and well-documented code, which is crucial for creating effective ML algorithms.

The ML engineer must be skilled in processing and analysing large data sets. Experience with data processing tools (e.g. Pandas, NumPy) and familiarity with SQL and NoSQL databases is key to preparing data for analysis and modelling.

In addition to skills directly related to ML, the candidate should have a good understanding of software engineering practices, including version control (Git), testing, containerisation (Docker) and CI/CD, which is important for implementing and maintaining production ML systems.

Effective communication and teamwork skills are key, as working on ML projects often requires collaboration with data analysts, software engineers and business stakeholders. The candidate should be able to clearly present complex ML concepts and the results of their work.