Principal Algorithm Scientist / Head of Recommendation Systems

Role Overview

We are seeking a Principal or Head-level expert with extensive experience designing and scaling recommendation systems from the ground up. This position suits someone passionate about data, advanced modeling, and production-grade coding who can take full ownership of recommender system stacks that serve millions of users.

Key Responsibilities

• Build complete recommendation engines including data pipelines, candidate generation, ranking, re-ranking, and serving infrastructures.
• Develop and enhance models such as collaborative filtering, content-based methods, two-tower architectures, sequential, and graph-based approaches.
• Design resilient strategies addressing cold-start, diversity, and balancing exploration versus exploitation.
• Transition prototypes to production-ready systems focusing on low latency, high scalability, and reliability.
• Implement real-time and batch inference pipelines, create feature stores, and set up monitoring systems for model drift and performance degradation.
• Collaborate closely with data and ML engineers on deploying and optimizing large-scale recommender platforms.
• Create sophisticated ranking models leveraging learning-to-rank methods, multi-task learning, and deep neural networks.
• Continuously improve recommendation relevance, user engagement, and conversion rates through experimentation.
• Define and oversee offline evaluation metrics such as NDCG, Recall@K, MRR, alongside online assessments like A/B tests and interleaving.
• Provide technical leadership by setting standards for coding, modeling, and evaluation in recommendations.
• Mentor a team of data scientists and ML engineers, driving technical design discussions and fostering best practices.
• Contribute hands-on coding and research approximately 70% of the time, while dedicating 30% to leadership duties.

Candidate Profile

• Minimum of 8 years experience in Data Science or Machine Learning, including at least 4 years focusing on production recommender systems.
• Proven record of delivering at least two large-scale recommendation engines serving live users with measurable improvements in click-through rates, conversion, retention, or engagement.
• Expertise in retrieval, ranking algorithms, embeddings, and sequential modeling techniques.
• Proficiency in Python programming, advanced SQL skills, and familiarity with machine learning frameworks like PyTorch or TensorFlow, along with Spark.
• Experience working with vector databases, approximate nearest neighbor (ANN) search tools such as FAISS, HNSW, and scalable model serving solutions like ScaNN.
• Strong understanding of A/B testing methodologies and the analysis of large-scale user behavioral data.
• Leadership experience guiding small teams (3-8 members) while maintaining an active hands-on role.

Desirable Qualifications

• Experience developing recommendation systems for marketplaces, e-commerce platforms, content providers, or OTT services.
• Knowledge in real-time personalization and session-based recommendation techniques.
• Familiarity with ranking infrastructures such as Vespa or Elasticsearch.