Role Overview
As an Intern Machine Learning Engineer on a 6-month contract, you will help build, test, and deploy ML-powered services on our provider data platform. This role focuses on strong software engineering, testing, and robust evaluation rather than novel model architectures. You will own features end-to-end by collaborating with stakeholders, implementing production-ready code, designing evaluation pipelines, and deploying services on Google Cloud Platform (GCP). This is a fully remote position.
Responsibilities
- Design, implement, and maintain ML-driven services and data workflows in Python.
- Apply software engineering best practices, including clean code, testing, code reviews, CI/CD, and observability.
- Build and maintain evaluation pipelines and metrics to measure model and system performance.
- Deploy and operate ML services on GCP, including tools such as Cloud Run, GKE, Cloud Functions, Pub/Sub, BigQuery, and Cloud Storage.
- Troubleshoot and improve existing ML services with a focus on reliability, latency, and correctness.
- Collaborate proactively with internal stakeholders across product, operations, engineering, and data teams.
Requirements
- Experience as a Software Engineer or Machine Learning Engineer.
- Strong proficiency in Python and experience building production services.
- Hands-on experience deploying and running workloads on Google Cloud Platform.
- Expertise in writing and debugging SQL queries.
- Strong foundation in software engineering fundamentals, including testing, debugging, Git, and CI/CD.
- Experience evaluating ML systems by defining metrics and building evaluation datasets.
- Ability to work independently and drive projects with limited supervision.
Skills
- Python
- Google Cloud Platform
- SQL
- CI/CD
- Git
Nice to Have
- Experience writing code in Java.
- Experience building or maintaining data pipelines or ETL jobs on GCP.
- Experience working with healthcare data or compliance requirements.
- Experience with experiment tracking tools such as MLflow or Weights & Biases.