How to apply for Senior Software Engineer, Scientific Computing

About KoBold Metals

KoBold Metals is a climate-tech company using AI to accelerate the discovery of critical minerals needed for the electric vehicle transition. By combining geoscience, machine learning, and large-scale data analysis, they aim to find new ore deposits faster and more efficiently than traditional methods. Working there means directly contributing to the fight against climate change while pushing the boundaries of scientific computing.

About the role

As a Senior Software Engineer, Scientific Computing, you'll build the foundational libraries and tools that enable KoBold's data scientists to predict ore locations. You'll architect scalable ML pipelines, optimize array-based computations, and create visualizations that make complex geological data accessible to domain experts. This role is pivotal in turning raw geodata into actionable insights that guide real-world exploration.

A typical day

Your day might start with a stand-up with data scientists to prioritize library features, followed by deep work implementing a new xarray-based algorithm for processing seismic data. In the afternoon, you might pair with a geologist to refine a visualization tool, then review a colleague's pull request on an ML pipeline component. You'll balance hands-on coding with mentorship, ensuring the team's codebase remains robust and scalable.

Who KoBold Metals is looking for

• 5–10 years of software engineering experience, with a strong focus on scientific computing and Python (especially numpy, xarray, and related libraries).
• Proven track record of building production-quality data processing systems and MLOps infrastructure, ideally for geospatial or measured scientific data.
• Deep understanding of machine learning concepts, from classical statistics to deep learning, and the ability to collaborate effectively with data scientists.
• Experience coaching engineers on best practices for robust, testable, and composable code, and a passion for writing clean, maintainable scientific software.

Tips for this application

• Highlight any experience with geospatial or earth science data (e.g., seismic, satellite, or sensor data) in your resume and cover letter—KoBold specifically values domain familiarity.
• Emphasize your work with xarray and numpy, and describe projects where you optimized array computations or built libraries used by other scientists.
• Showcase MLOps experience by detailing pipelines you've built, including how you ensured reproducibility, scalability, and monitoring.
• Mention any contributions to open-source scientific Python libraries or talks/papers you've given—KoBold values community engagement and technical leadership.
• Tailor your resume to include keywords like 'mineral exploration', 'ore concentration prediction', and 'scientific data visualization' to align with the job description.

What to cover in your cover letter

["Your specific experience with Python array-based packages (xarray, numpy) and how you've used them to solve complex scientific problems.", 'A concrete example of building ML infrastructure that accelerated model development or deployment for a scientific team.', 'Your ability to bridge the gap between software engineering and geoscience, perhaps through previous collaborations with domain experts.', "Your motivation to apply your skills to climate change and mineral exploration, showing alignment with KoBold's mission."]

Research before applying

• Learn about KoBold's specific exploration projects and the types of data they work with (e.g., geochemical, geophysical, remote sensing).
• Understand the basics of mineral exploration and ore deposit formation to speak the language of geoscientists.
• Read KoBold's technical blog posts or publications to understand their tech stack and approach to ML in exploration.
• Research the founders and leadership team, especially their backgrounds in AI, geoscience, and climate tech.

Likely interview topics

Based on the job description, expect questions about:

• How you would design a library for efficient processing of large-scale geospatial data using xarray and numpy.
• Your approach to building MLOps pipelines that support rapid experimentation while maintaining production reliability.
• A time you coached a team member on writing testable and composable code, and the outcome.
• How you would visualize complex geological data for non-technical domain experts to aid decision-making.
• Your experience with measured scientific data and how you handle noise, uncertainty, and missing values.

Common mistakes to avoid

• Focusing only on generic software engineering skills without demonstrating experience in scientific computing or ML.
• Neglecting to mention specific Python libraries like xarray or numpy, which are explicitly required.
• Failing to show enthusiasm for the mission or understanding of how mineral exploration contributes to electrification.