Okko interview series: recruiter call for recommendations/search teams, Python/SQL coding, ML oral interview on math and systems basics, then ML System Design for content-aware similar-items recommendations.
Последовательность опубликованных собеседований в рамках одного процесса.
Recruiter explains recommendation and search team scope, personalized homepage projects, remote format and Python/SQL -> ML -> ML System Design interview series.
Python internals, argument passing, FastStatistics, embedding-code review, SQL LAG and Precision/Recall@K from movie recommendations.
Probability spaces, vector spaces, CLT, systems basics around REST/gRPC and database indexes, plus product context for video understanding and agents.
System design case for cold-start similar items: multimodal embeddings, ANN retrieval, feedback-trained relevance, hard negatives, offline metrics and A/B validation.
