Overstory is looking to solve the climate crisis by enabling a resilient electrical grid. The problem stems from extreme weather, aging infrastructure, and growing wildfire risks putting the grid under pressure. A leading cause of these issues is trees and brush contacting power lines. Overstory uses AI and satellite imagery to identify and prioritize vegetation risks to prevent outages, reduce wildfire risks, and accelerate the transition to a safer, more resilient grid.
Requirements
- Designing and building production-grade ML pipelines and systems (8+ years, 5+ years considered)
- Strong knowledge of experiment tracking
- Strong knowledge of model deployment strategies
- Strong knowledge of data versioning
- Strong knowledge of monitoring
- Experience with ML infrastructure tools (e.g. MLflow, Kubeflow, Airflow, feature stores, model registries)
- Familiarity with GCP and VertexAI preferred
Responsibilities
- Design, build, and maintain automated pipelines for training, testing, and deploying ML models
- Design, build, and maintain experiment tracking systems for performance metrics, data and model versioning, and documentation
- Design, build, and maintain processes and systems for the full model lifecycle, including registries, release and rollback strategies, and scalable model serving
- Design, build, and maintain monitoring and alerting for prediction quality, system health, and cost optimization
- Advocate for a balance between MLOps best practices and quick slices of value
- Align technical solutions with customer needs in collaborating with both engineering and product
- Ensure our MLOps systems support regulatory, privacy, and security requirements
Other
- You love working in a remote-first, fast-moving environment where collaboration and adaptability are essential.
- Strong communication skills and ability to align technical solutions with business goals.
- Comfortable making architectural decisions and balancing best practices with practical trade-offs.
- Experience in remote-first or globally distributed teams.
- Background in geospatial or spatio-temporal data processing.
- Prior work on real-time prediction systems or active-learning loops.
- Knowledge of regulatory, privacy, or security considerations in ML.
- Experience optimizing cloud infrastructure costs for ML workloads.
- Familiarity with Overstory’s mission domains (e.g. satellite imagery, forestry, utilities)
- Candidates must be living and working in one of these countries: United States, the Netherlands, United Kingdom, Ireland, Estonia, Portugal, France, Sweden, Denmark, Switzerland, and Canada.
