- Published on
- Posted 5 days ago
Machine Learning Engineer
- Authors
- company
- MediaLab
- Company
- medialab.la
Job Details
Role: Machine Learning Engineer
Location: Remote - LATAM (Argentina / Mexico)
Job Type: Contractor (Remote)
Department: Data
About MediaLab
MediaLab is a media & technology company focused on acquiring and growing global brands. We combine private equity, holding company, and operating entity strategies to create a unique company structure. Headquartered in Santa Monica, California, MediaLab also has growing teams in New York and across Latin America.
About the Role
As a Machine Learning Engineer at MediaLab, you will ensure that the foundation for our machine learning models is solid and scalable. Working closely with the data science and business intelligence teams, you'll be responsible for optimizing and productionalizing ML models, ensuring they run efficiently at scale. You will handle model deployment, build ML workflows, and maintain the retraining of models in production.
Responsibilities
- ML Workflow Development: Implement a framework for model development and deployment.
- Data Sourcing & Cleansing: Work with data science and BI teams to gather, clean, and prepare data as model inputs.
- Data Integration: Coordinate with other teams to access necessary data sources and follow through to ensure data acquisition.
- Model Optimization: Optimize models to run efficiently and cost-effectively at scale.
- Productionalize Models: Ensure models are deployed with accessible outputs, working with data engineers and software engineers.
- API Integration: Expose model outputs to applications via database connectors, RESTful APIs, or other methods.
- Model Retraining: Manage the ongoing scheduled retraining of models, validating outputs with data science.
Requirements
- Experience:
- 3-5 years of professional experience in ML Engineering, ML Ops, or as a Data Engineer with significant exposure to ML model deployment.
- Technical Skills:
- Advanced Python skills and experience with ML workflows or **CI/CD