Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
You will formulate models of user preference, similarity, and engagement for music discovery. You will design learning systems from interaction data, build and deploy scalable recommendation and ranking models, define optimization metrics, run large-scale experiments and causal analyses, and help set the technical direction for personalization systems.
Requirements
- Applied mathematics, statistics, machine learning, or related quantitative-field background
- PhD or equivalent experience
- Experience designing probabilistic models, optimization-based systems, representation-learning models, or graph-based methods
- Python
- PyTorch
- Experience with implicit feedback, ranking losses, bandits, or reinforcement-learning-adjacent methods
- Ability to evaluate tradeoffs among model quality, scalability, and system constraints
- Eligibility to work in the United States
Responsibilities
- Develop mathematical models of user preference, similarity, and engagement for music discovery
- Design learning systems that infer user taste from sparse, noisy, and evolving interaction data
- Build and deploy scalable real-time recommendation and ranking models
- Translate relevance, novelty, diversity, and long-term satisfaction objectives into measurable metrics and optimized systems
- Run large-scale experiments and causal analyses to evaluate model behavior and product impact
- Define the technical direction of personalization systems with product and research leadership
Benefits
- Company equity package
- 401(k) with 3% employer match and Roth 401(k)
- Medical, dental, and vision insurance with HSA and FSA options
- 11 paid holidays
- Unlimited PTO and sick time
- 16 weeks of paid parental leave
- Creative education stipend
- Commuter allowance
- In-office lunch five days per week