ML engineer
Also known as: machine learning engineer, MLE
ML engineers have been a fixture of tech teams since around 2016. They specialize in the pipeline from raw data to deployed model: data preprocessing, model architecture choices, training runs, evaluation, and serving infrastructure. Unlike data scientists, who often focus on analysis and experimentation, ML engineers prioritize production readiness, scale, and reliability.
The rise of large language models and foundation models has shifted some of the work. Many ML engineers now spend more time fine-tuning pre-trained models, building evaluation harnesses (test suites for measuring model behavior), and wiring models into applications via APIs than they do training models from scratch. The tooling has also matured, with MLOps frameworks, experiment trackers, and model registries handling much of the infrastructure work.
In practice, job postings for ML engineer and AI engineer now overlap significantly. The distinction, where it exists, is that ML engineers tend to be closer to model internals: training pipelines, data quality, and optimization. AI engineers tend to work at the application layer, connecting model APIs to products. Both roles are genuinely in demand, and many builders move fluidly between them.