Design Employee-to-Employee Distance
Company: Glean
Role: Machine Learning Engineer
Category: ML System Design
Difficulty: medium
Interview Round: Onsite
Design an employee-to-employee distance system for a company.
The system should take two employees as input and return a meaningful "distance" or similarity score between them. The distance may be used for products such as people search, internal networking recommendations, collaboration discovery, org insights, or onboarding suggestions.
Discuss:
- What "distance" should mean and how it should vary by use case.
- What employee data you would use, such as org chart, team, manager chain, office location, projects, skills, documents, communication metadata, and collaboration history.
- How you would model the relationship between employees.
- How to serve low-latency pairwise distance queries.
- How to evaluate quality, freshness, privacy, fairness, and abuse risks.
Quick Answer: This question evaluates competencies in ML system design, similarity and representation modeling, feature and data engineering, scalable low-latency serving architectures, and privacy, fairness, and evaluation practices.