Machine Learning Engineer

Roche•Mississauga, ON

About The Position

At Roche, we are committed to delivering greater benefits to our patients by applying digital, data, machine learning, and AI capabilities to real-world operational challenges. The ML/MLOps Engineer is a hands-on technical contributor within the MLE/DE Cluster, focused on building, deploying, monitoring, and improving machine learning solutions for Pharma Technical Operations. This role contributes to the end-to-end machine learning lifecycle, including data preparation, feature engineering, model training support, experiment tracking, model packaging, deployment pipelines, model serving, monitoring, and continuous improvement. The role works closely with data scientists, AI engineers, software engineers, data engineers, process experts, IT, quality, and business stakeholders to help deliver reliable, scalable, and compliant ML-enabled solutions in manufacturing, quality, supply chain, and technical operations environments. The role also contributes to modern ML engineering capabilities, including data science-driven evaluation frameworks, reusable ML evaluation harnesses, model lifecycle management, reproducible experimentation, production monitoring, and MLOps practices that help move solutions from prototype to reliable production use.

Requirements

Bachelor's degree in computer science, data science, engineering, mathematics, statistics, or a related field; or equivalent practical experience.
Relevant experience building software, data, analytics, machine learning, or MLOps solutions.
Strong hands-on Python skills and understanding of software engineering practices.
Experience with Git, APIs, testing, documentation, and collaborative development.
Exposure to machine learning workflows, including data preparation, feature engineering, model training, evaluation, deployment, and monitoring.
Exposure to MLOps tools and practices such as MLflow, model registries, experiment tracking, CI/CD, Docker, automated testing, and model monitoring.
Experience working with SQL and data pipelines.
Exposure to cloud platforms, preferably AWS, and containerized deployment approaches.
Ability to work in cross-functional teams and communicate technical topics clearly.
Interest in building production-grade ML systems rather than one-off prototypes.

Nice To Haves

Experience in industrial, manufacturing, pharmaceutical, biotechnology, quality, supply chain, or regulated environments is an advantage.

Responsibilities

Build and maintain components of machine learning pipelines, including data ingestion, preprocessing, feature generation, model training support, validation, packaging, and deployment.
Support the deployment of ML models into cloud, hybrid, or application environments using reliable, repeatable, and automated deployment practices.
Contribute to model serving components, APIs, batch scoring workflows, and integration patterns that make ML outputs available to business applications and operational users.
Use experiment tracking tools and structured documentation to support reproducibility, comparison of model versions, and evidence-based model improvement.
Contribute to data science-driven evaluation frameworks, including benchmark datasets, validation checks, model performance metrics, regression testing, and comparison of model outputs.
Build and maintain components of reusable evaluation harnesses that allow repeatable testing of models, data pipelines, features, scoring logic, and model performance across versions.
Support monitoring of deployed ML models, including performance, data quality, drift, latency, reliability, usage, and operational failure signals.
Help investigate issues in ML pipelines and deployed models, support incident resolution, and contribute to continuous improvement of production ML systems.
Apply solid Python development practices, Git-based collaboration, automated testing, code reviews, documentation, and maintainable code design.
Apply foundational MLOps concepts such as model versioning, model registry usage, experiment tracking, CI/CD, automated testing, deployment workflows, and monitoring.
Work with SQL, structured and semi-structured data, operational data sources, time-series data, and data quality checks needed for ML solutions.
Understand common ML workflows, including supervised learning, validation, feature engineering, model evaluation, model packaging, and model inference.
Work with cloud and hybrid environments, preferably AWS, as well as Docker, CI/CD pipelines, infrastructure automation concepts, and runtime monitoring.
Build ML components with reliability, traceability, reproducibility, observability, and maintainability in mind.
Follow Roche standards, quality expectations, security requirements, data privacy expectations, and responsible AI practices, especially where ML solutions may support regulated environments.
Work with data scientists, AI engineers, software engineers, data engineers, IT, quality, process experts, and business stakeholders to understand use cases and support technical delivery.
Help translate data science prototypes into robust, maintainable, and deployable ML solutions.
Actively learn new MLOps, ML engineering, cloud, and software engineering methods and share relevant insights with the MLE/DE Cluster.
Be curious, open to feedback, and willing to develop deeper expertise in machine learning, manufacturing, quality, and regulated production environments.
Support effective and responsible adoption of machine learning and AI within Pharma Technical Operations.