Software Engineering Internship, AI Planet

Persons in these roles are expected to work from our offices in Seattle. On-site requirements vary based on position and team. If you have questions about on-site work arrangements for this role, please ask your recruiter.
* Compensation dependent on current degree enrollment.
Engineering Intern, Undergrad: $94,320 (Annually)
Engineering Intern, Masters: $106,110 (Annually)
Engineering Intern, PhD: $140,000 (Annually)

Who You Are:

We are seeking a motivated and passionate Software Engineering Intern to join the Skylight Engineering Team for the summer. This role offers an exciting opportunity to work on impactful projects, learn from experienced engineers, and contribute to protecting marine life.

Ai2 Internship Information:

• Duration: 12 weeks
• Start date:
  
  Flexible
• Candidates:
  An undergraduate student with strong software engineering fundamentals and a passion for building products with impact.

To see more about our current, future, and past interns, check out our internship page on our website!

Who We Are:

Ai2's AI for the Planet initiative brings together programs like Skylight and Olmo Earth to apply cutting-edge AI toward solving urgent environmental and humanitarian challenges. From monitoring illegal fishing and protecting marine ecosystems to mapping crops, tracking wildfires, supporting disaster recovery, and strengthening food security, our teams build open, scalable systems. These systems turn global Earth-observation data into actionable insights.

By combining advanced machine learning with open infrastructure and a deep commitment to public impact, we empower organizations worldwide to better understand and protect our planet's most vital systems.

Read more about Olmo Earth, and Skylight on our websites.

At Ai2, we are dedicated to using AI to solve some of the world's biggest problems. Our mission-driven team values collaboration, impact, and openness. As an intern, you'll be part of an inclusive and supportive environment where your ideas and contributions truly matter.

Your Next Challenge:

• Collaborate with the engineering team to develop, test, and deploy software tools that support our mission.
• Work on and ship production code for real-world projects involving data pipelines, web applications, or satellite imagery processing.
• Gain hands-on experience with technologies such as Typescript, Python, React, Kafka, Mapbox, Elastic Search, Postgres, Mongo
  
  DB, Kubernetes, and cloud platforms like Google Cloud.
• Participate in brainstorming sessions, technical discussions, and code reviews.
• Contribute to the development of tools or features that extend our platform's ability to deliver high-impact insights across domains such as marine conservation, land use, climate resilience, and disaster recovery.

What You'll Need:

• Undergraduate or Masters student pursuing a degree in Computer Science or a related field.
• Passion for software engineering and interest in environmental conservation.
• A strong foundation in any modern programming language.
• Curiosity and eagerness to learn new technologies and tackle technical challenges.
• Strong communication skills and ability to work in a collaborative team environment.
• Interest in the full product development process, including engaging with users to understand their needs and gather feedback to improve products.

• Real-world experience in software engineering and exposure to state-of-the-art AI tools.
• Mentorship from experienced professionals who are leaders in their fields.
• Insight into how technology is used to address critical environmental challenges.
• A chance to make a meaningful impact, and use technology for good, tackling some of the planet's most pressing environmental and humanitarian challenges.

Projects:

To give you a better idea of the kinds of projects you'd work on, here are some examples of our previous and current projects:

• Multisensor Satellite Detection Pipelines - What began as a Sentinel-2 vessel-detection system has evolved into a planetary-scale inference pipeline spanning Sentinel-1, Sentinel-2, and Landsat imagery. Our distributed processing stack downloads new satellite scenes as soon as they're available, runs fine-tuned computer-vision models, and makes results visible to users within hours. Building this system meant solving hard problems in cloud-scale parallelism, planetary datasets, and GPU optimization.
  
  (Read more in our blog post here)
• Geospatial Annotation Tool - We built a browser-based annotation platform that lets users create training datasets across multiple imagery types and time steps. It's used by conservation scientists and NGOs to label everything from croplands to mangroves, replacing what was once a fragmented or prohibitively expensive process.
• Model Fine-Tuning Platform - The Olmo Earth Platform allows users with no ML background to fine-tune our foundation models to their specific domain - whether that's wildfire fuel mapping, crop classification, or flood detection. Engineers collaborated closely with researchers to design the…