Mozilla.ai Builders in Residence

Mozilla.ai Builders in Residence

• Location: Remote (Europe, USA, Canada)
• Type: Contractor; Full-Time
• Expected Start Date: June 2025

Position overview

Are you looking to build a future where open source AI is the de facto standard?  Where ML expertise is valued but is not a prerequisite for owning your own AI workflows and the data generated by them?  Where you can build, experiment, and deploy under your own rules, without needing to call out to a third party?  Come spend some time with us at Mozilla.ai, working with our team of mission-driven ML and engineering experts on creating that future.

We’re looking for practical, application-driven researchers to join our team for a fixed period of time.  We’re setting a baseline of 3 months to start, with the option to extend depending on project scope and your circumstances.  This role is best suited to a PhD student seeking an internship/project; a recent PhD graduate or an Applied Scientist working in tech companies interested in a short-term project.  You have opinions about what we should build and open up to the community (via code, documentation, and whitepapers) to promote the future you want.

We’re running this program for the first time at Mozilla.ai, but something similar led by our CEO at his last startup resulted in work in Nature Machine Intelligence, ICLR, the most-cited counterfactual explanations survey, AIES (1, 2, 3), NeurIPS, AAAI, etc., as well as follow-on positions at CMU, Cornell, MSR, NIST, OpenAI, Stanford CS, Stanford Health Policy, Stanford Law, and more. We’d love to have you join to help push our mission forward while accelerating your career.

Sample focus areas

• Promote and measure agent effectiveness and safety: explore evaluation and validation methods for agent and multi-agent systems.
• Understand the Pareto frontier of the complexity of multi-agent interactions with the expressiveness of those transaction networks. More complex networks means more generality, but more computation and more data needed to evaluate that system.
• Analyze how data and control flow through compositions of actions or workflows, and use that to provide guaranteed guardrails.
• Expand our understanding of meta-learning for recommender systems, especially recommender systems over complex, combinatorial spaces.
• Characterize efficient multi-agent communication: that could mean prompt minimization, agent-to-agent language specialization and optimization, or something else.
• Create new interfaces and new user experiences for humans and multi-agent systems to collaborate, and understand what a future of human-AI interaction means for knowledge search and generation.
• BYO pitch!  We’re all optimistic tech nerds here, so feel free to pitch your own project.  We are emphatically not a blue-sky R&D organization, though, so we’ll need to ensure alignment with our mission and our focus – democratizing open source AI – before giving the go-ahead.

What you will do

• Collaborate with our team.  We are a highly-technical and engineering-heavy team focused on building, and supporting, a future where open source AI is the clear choice.
• Contribute to open source codebases, ensuring high-quality, interoperable, and accessible software.
• Work toward other forms of publishable artifacts such as academic whitepapers and open datasets.  We can cater this to your interests and your project.

Qualifications

• Pursuing a PhD degree in ML/AI, Computer Science/Engineering, Mathematics, Statistics, or related disciplines; or equivalent experience in applied R&D
• Strong fundamentals in machine learning and AI
• Lead-author publications in top ML/AI conferences and ML-adjacent venues (e.g., ICLR, ICML, NeurIPS, AAMAS, COLM, AIES, FAccT, RLC, CHI, CSCW, ACL, EMNLP, etc)
• Experience developing or interest in multi-agent systems and/or federated learning
• Strong development skills in Python
• Experience using core machine learning frameworks such as PyTorch, JAX, HF Transformers, Gym/Gymnasium, etc
• Excellent communication, teamwork and a problem solving, results-oriented attitude

What you bring

• Startup Mindset: Take ambiguous problems with creativity and rigor, prioritize tasks efficiently in a fast-paced environment, and bring solutions from ideation to validation quickly.
• Open-Source: expertise using and developing open-source tools and libraries, and fostering a culture of open collaboration.
• Continuous Learning: Stay up-to-date with the latest research and industry advancements in machine learning, generative AI, and open-source technologies, applying new ideas to improve projects and products.
• Creativity: We’re building in a broad open space in which the greatest opportunities may not have even been identified yet.
• Opinions: As you know, we’re working in a very fast-moving space, and the right answers to questions aren’t always obvious.  We need to make bets, which means being informed and confident enough to know when to say yes and when to say no to ideas.

Why us 

We are more than just a company; we are a community of like-minded individuals driven by a shared passion for creating positive change in society through AI solutions.

1. Purpose-Driven Mission: we are a mission-driven early stage company. If you are passionate about the transformative potential of AI and committed to ensure AI solutions that are trustworthy and responsible.
2. Innovation & Impact: cutting-edge AI projects that have a real impact on people's lives.
3. Collaborative Culture: Our team is distributed across different countries, fostering a collaborative and inclusive culture where everyone's input is valued. We make sure to meet several times a year to work together in a place in the world defined in advance.
4. Remote work: We are a 100% remote team, distributed around the world. 

We are committed to building a diverse and inclusive team. We encourage applications from individuals of all backgrounds, beliefs, and identities.