Enhance Reddit's recommendation systems by developing highly expressive, multi-entity embeddings to improve personalization and relevance across the platform.
Requirements
- 5+ years of experience in machine learning engineering, with a strong focus on recommendation systems, representation learning, and deep learning.
- Hands-on experience with Graph Neural Networks (GNNs), collaborative filtering, and large-scale embeddings.
- Proficiency in Python and experience with ML frameworks such as PyTorch Geometric (PyG), Deep Graph Library (DGL), TensorFlow, or JAX.
- Strong understanding of graph theory, network science, and representation learning techniques.
- Experience building distributed training and inference systems using ML infrastructure components (data parallelism, model pruning, inference optimization, etc.).
- Ability to work in a fast-paced environment, balancing innovation with high-quality production deployment.
Responsibilities
- Design and implement scalable, high-performance machine learning models using Graph Neural Networks (GNNs), transformers, and knowledge graph approaches.
- Develop and optimize large-scale embedding generation pipelines for Reddit’s recommendation systems.
- Collaborate with ML infrastructure teams to enable efficient distributed training (multi-GPU, model/data parallelism) and low-latency serving.
- Drive feature engineering efforts, identifying and curating expressive raw data to enhance model effectiveness.
- Monitor, evaluate, and improve model performance using A/B testing, offline metrics, and real-time feedback loops.
- Stay up-to-date with the latest research in GNNs, transformers, and representation learning, bringing new ideas into production.
- Participate in code reviews, mentor junior engineers, and contribute to technical decision-making.
Other
- Work closely with cross-functional teams (Ads, Feed Ranking, Content Understanding) to integrate embeddings into various personalization and ranking systems.
- Strong communication skills and the ability to collaborate cross-functionally with engineers, researchers, and product teams.
- Flexible Vacation & Reddit Global Days off
- Generous paid Parental Leave
- Paid Volunteer time off
- Comprehensive Healthcare Benefits and Income Replacement Programs
- 401k Match
- Family Planning Support
- Gender-Affirming Care
- Mental Health & Coaching Benefits
