How to apply for Staff Applied Scientist - Hyperpod

About KoBold Metals

KoBold Metals is an AI-native mineral exploration company using artificial intelligence to find critical minerals like copper, lithium, and cobalt. Its first discovery, Mingomba in Zambia, is a high-grade copper deposit found at materially less than industry average cost. The company combines data science, engineering, and geology to make mineral exploration a repeatable science.

About the role

As a staff applied scientist on the Hyperpod team, you will build and scale systems that support KoBold's subsurface prediction workflows across exploration projects. This role sits at the intersection of applied machine learning, data infrastructure, and scientific computing, directly supporting the company's goal of faster and cheaper mineral discovery. Your work will help turn exploration data into testable predictions used on real projects like Mingomba.

A typical day

A typical day might involve reviewing model training runs on large geoscience datasets, debugging data pipeline issues with engineers, and meeting with geologists to refine prediction targets. You would also spend time designing experiments or infrastructure improvements for the Hyperpod platform. Exact routines depend on team structure, so ask about daily collaboration patterns during interviews.

Who KoBold Metals is looking for

• Has senior or staff-level experience building applied machine learning systems, ideally in scientific computing, geoscience, or a related technical domain.
• Can design and operate large-scale data pipelines and compute infrastructure, with familiarity with distributed training or high-performance computing environments.
• Comfortable working with messy, multi-modal scientific data such as geophysical, geochemical, or drilling data.
• Collaborates well with geologists, data scientists, and software engineers, and can translate domain problems into technical solutions.

Tips for this application

• Read KoBold's public materials on Mingomba and their exploration approach before writing anything, then reference specific technical challenges from those materials in your application.
• Highlight any experience with scientific data pipelines, distributed compute, or ML infrastructure, since the Hyperpod role likely involves scaling model training and data processing.
• Name concrete examples of applied ML systems you have shipped, including the scale of data and compute involved, because KoBold values repeatable science over research prototypes.
• If you have worked with geoscience, mining, or subsurface data, state it clearly and early, even if it was a smaller part of a past role.
• Ask directly in your cover letter or interview how the Hyperpod team interacts with KoBold's exploration projects, since the job description does not specify team structure or daily workflows.

What to cover in your cover letter

Explain why AI-native mineral exploration matters to you using facts about KoBold's work, not general climate statements. Describe a specific applied ML or data infrastructure project you led and the measurable outcome. Connect your technical skills to the likely demands of Hyperpod, such as large-scale compute, model training, or scientific data handling. Show that you understand KoBold's model of integrating data scientists, engineers, and geologists on real exploration projects.

Research before applying

• KoBold's public announcements and technical blog posts about Mingomba and their exploration methodology.
• The backgrounds of KoBold's leadership and technical team, especially anyone associated with Hyperpod or subsurface prediction.
• Industry context on why mineral discovery rates have fallen and how AI is being applied to exploration, to understand the problem space.
• Any open-source tools, papers, or talks from KoBold that mention Hyperpod or their data infrastructure stack.

Likely interview topics

Based on the job description, expect questions about:

• How you would design a scalable pipeline for training and deploying models on multi-modal geoscience data.
• Your experience with distributed compute frameworks and how you handle resource scheduling or fault tolerance in large jobs.
• A time you worked with domain experts to turn a scientific problem into a machine learning task.
• How you evaluate model performance when ground truth is sparse or expensive to obtain, as in mineral exploration.
• Why you want to apply staff-level applied science to mineral exploration rather than a more traditional tech domain.

Common mistakes to avoid

• Submitting a generic applied scientist resume that does not mention scientific data, geoscience, or large-scale compute, which signals a lack of fit for this specific role.
• Writing a cover letter focused only on climate change without connecting to KoBold's technical approach or the Hyperpod team's likely work.
• Claiming expertise in mineral exploration without evidence, when the job description shows KoBold values real domain collaboration over buzzwords.