AE Studio
Alignment Scientist/Engineer
About this role
AE Studio, an ML consultancy with a dedicated AI safety research arm, seeks an experienced alignment scientist or engineer to join their 25-person research team. You'll work on cutting-edge alignment challenges including interpretability, model safety, and neuroscience-inspired approaches, partnering with top external researchers while maintaining significant autonomy over your work.
What you'll do
- Conduct independent and collaborative ML/alignment research on novel safety techniques
- Design and execute experiments to test alignment interventions and hypotheses
- Develop and evaluate deep learning models using modern tools and APIs
- Communicate research findings to technical and non-technical audiences
- Partner with external researchers from academia and industry on alignment problems
- Guide research direction alongside collaborators in a fast-paced environment
What they're looking for
- Machine learning research and/or AI alignment expertise
- Advanced Python and PyTorch proficiency
- Deep learning model development and evaluation
- Problem decomposition and experimental design
- REST APIs and ML client libraries
- Technical communication and writing
- Self-directed research capability
- Software engineering best practices (bonus)
Benefits
- Remote-first with flexible scheduling
- Health insurance
- Free lunches at LA office
- Weekly knowledge-sharing talks
- Annual team retreat
- Performance-based raises and employee equity program
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
AE Studio
aestudio is an ML consultancy with a dedicated AI safety research arm focused on cutting-edge alignment challenges including interpretability, model safety, and neuroscience-inspired approaches. The company is hiring experienced alignment scientists and engineers to join their research team and work on these critical problems.
View all jobs at AE StudioLikely interview questions
- Walk us through a recent ML or alignment research project you've led—how did you decompose the problem and what experiments validated your approach?
- Describe your experience with PyTorch and the modern deep learning stack. What's a complex model or training challenge you've debugged?