ML Engineer (Recommendations)

Darwin • Palo Alto, CA

19h

About The Position

We’re looking for an Applied ML Engineer to design, evaluate, and scale recommendation and ranking systems that power how content, ads, and interactive experiences are selected and surfaced in real time. This role focuses on decision-making systems, with an emphasis on relevance, personalization, exploration, and long-term user value. You’ll work at the intersection of machine learning, product, and systems engineering: experimenting with modern recommender techniques, grounding them in real user behavior, and shipping models that perform reliably at scale. The work is hands-on, impact-driven, and tightly coupled to product outcomes rather than academic benchmarks alone.

Requirements

• Strong background in machine learning with a focus on recommender systems, ranking, or decision-making models.
• Experience with techniques such as matrix factorization, embeddings, sequence models, bandits, or reinforcement learning.
• Ability to implement and adapt methods from recent research or production-grade open-source systems.
• Proficiency in PyTorch (or equivalent) and experience training, evaluating, and deploying ML models.
• Comfort working with large-scale behavioral data and noisy, implicit feedback signals.
• Product intuition: ability to connect model improvements to real user and business outcomes.

Nice To Haves

• Experience with large-scale recommender systems in production environments.
• Familiarity with exploration–exploitation tradeoffs and long-horizon optimization.
• Publications or applied research in recommender systems, information retrieval, or applied ML.

Responsibilities

• Design and iterate on recommendation, ranking, and retrieval systems for content, creators, and ad placements.
• Track and evaluate advances in recommender systems, including representation learning, sequential models, bandits, and RL-based approaches.
• Build and test candidate generation, ranking, and re-ranking pipelines using offline and online signals.
• Develop experimentation frameworks to measure tradeoffs across relevance, diversity, exploration, latency, and long-term engagement.
• Work closely with product and engineering to translate user intent and business goals into modeling objectives.
• Improve feedback loops using implicit signals (clicks, dwell time, skips, conversions) and sparse explicit feedback.
• Optimize systems for real-time inference, scalability, and robustness under non-stationary user behavior.
• Document and communicate learnings internally, helping shape Darwin’s recommendation strategy and technical direction.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed