Design a system for LinkedIn Skills
Company: LinkedIn
Role: Machine Learning Engineer
Category: ML System Design
Difficulty: medium
Interview Round: Onsite
Design an ML system for “LinkedIn Skills”. The system should infer and/or recommend skills for members, and support downstream use cases like search/ranking, recruiter matching, and personalization.
Cover:
1) What are the core user-facing outputs? (e.g., skill inference, skill recommendation, skill confidence)
2) Data sources and labeling strategy.
3) Feature engineering and modeling approach (including how you would use LLMs).
4) Online serving architecture and monitoring.
5) Evaluation (offline + online) and how to handle quality issues (hallucinated skills, stale skills, spam).
Overview: This question evaluates a machine learning engineer's ability to design an end-to-end ML system for skill inference and recommendation on a professional networking platform, testing competencies in data sourcing and labeling, feature engineering, model selection (including LLM integration), online serving, monitoring, and quality control.
Read the full LinkedIn Machine Learning Engineer interview experience this question came from