Recommendation Systems Engineer Job Description Template

The recommendation systems engineer builds the models and infrastructure that decide what a user sees next. The work combines candidate generation, ranking and serving under strict latency budgets, and it carries a particular evaluation challenge, since offline metrics correlate imperfectly with what actually happens when real users are shown the results.

Typical Duties and Responsibilities

• Build candidate generation and ranking models
• Design feature pipelines with consistent offline and online computation
• Serve recommendations within tight latency budgets
• Design and run online experiments to measure real impact
• Address cold start for new users and new items
• Balance relevance against diversity, freshness and business objectives
• Monitor for feedback loops and popularity bias
• Investigate discrepancies between offline and online results
• Optimize infrastructure cost at high request volume
• Work with product on what the system should optimize for

Education

• Bachelor’s or Master’s degree in computer science, machine learning or a related field

Required Skills and Experience

• 3+ years building recommendation or ranking systems
• Strong Python and machine learning engineering skills
• Understanding of retrieval and ranking architectures
• Experience with online experimentation and its interpretation
• Ability to serve models under low latency constraints
• Awareness of feedback loops and bias in recommender systems
• Experience with large scale feature pipelines
• Judgment about competing objectives

Preferred Qualifications

• Experience with embedding based retrieval at scale
• Exposure to real time or streaming feature computation