How to apply for Staff Machine Learning Engineer - Wildfire

About Overstory

Overstory uses satellite imagery and AI to find vegetation that threatens power lines. The company works with utilities in the Americas and Europe to reduce wildfire risk and outages. This is a chance to work on grid resilience and climate adaptation with a remote team.

About the role

This staff machine learning engineer role focuses on wildfire risk products. You will build and improve ML models that turn satellite data into vegetation risk analysis for utilities. The work directly affects where utilities trim trees and inspect lines, which can prevent fires and outages.

A typical day

A typical day might include reviewing satellite image data, writing code to improve a vegetation detection model, and meeting with product or engineering teammates to plan the next model iteration. You may also analyze model errors, prepare results for utility customers, or discuss data quality issues with the data team. Because the team is remote, much of the work happens asynchronously with some scheduled calls.

Who Overstory is looking for

• Has senior or staff-level experience building and deploying machine learning models in production.
• Has worked with geospatial data, satellite imagery, or remote sensing. If not, can show strong transferable experience with large image datasets.
• Can design end-to-end ML systems, including data pipelines, model training, evaluation, and monitoring.
• Is comfortable working remotely with a distributed team and communicating with utility customers or product teams.
• Has interest in climate adaptation, wildfire risk, or grid reliability. Prior utility experience is not required but is a plus.

Tips for this application

• In your resume, list specific ML projects where you worked with imagery, geospatial data, or large-scale sensor data. Include model types, data size, and business impact.
• Mention any experience with remote sensing libraries or tools such as GDAL, rasterio, or Google Earth Engine. If you have none, describe similar large-image or time-series work.
• Show that you understand the utility problem: vegetation management, wildfire risk, and grid reliability. Use the company website and any public case studies to name specific use cases.
• Since the role is remote in the US, state your location and time zone clearly. Mention any experience working with teams across the Americas and Europe.
• Avoid generic climate statements. Instead, connect your ML work to reducing outages or wildfire risk in concrete terms.

What to cover in your cover letter

['Why you want to apply ML to satellite imagery for wildfire risk, not just climate in general.', 'A specific example of an ML system you built that handled large geospatial or image datasets and delivered measurable results.', 'How you have worked with cross-functional teams, such as product, engineering, or customers, to turn model output into action.', 'Your experience with remote work and distributed teams, especially across US and European time zones.']

Research before applying

• Read Overstory's website and any public blog posts or case studies about how their vegetation risk product works.
• Look up the company's utility partners and the regions they serve. Understand why wildfire risk and grid reliability matter in those areas.
• Find public information about satellite imagery sources and ML methods used for vegetation detection. This will help you speak the same language in interviews.
• Check the LinkedIn profiles of current Overstory ML engineers to understand their backgrounds and the tools they use.

Likely interview topics

Based on the job description, expect questions about:

• How you would design an ML pipeline to detect vegetation encroachment from satellite imagery at scale.
• How you evaluate model performance when ground truth is limited or expensive to collect.
• A time you deployed a model to production and monitored it for drift or failure.
• How you would prioritize model improvements when the business goal is reducing wildfire risk for utilities.
• How you explain model limitations and uncertainty to non-technical stakeholders, such as utility customers.

Common mistakes to avoid

• Submitting a generic ML resume that does not mention geospatial data, satellite imagery, or climate applications.
• Claiming to care about climate change without connecting it to the specific problem of wildfire risk and grid reliability.
• Ignoring the remote and distributed nature of the team. Not stating your time zone or experience working across regions.
• Focusing only on model accuracy and ignoring deployment, monitoring, or how model output drives business decisions.