← Google Interview InsightsMachine Learning Engineer

Google·Machine Learning Engineer·Onsite - System Design / Architecture·Senior

SeniorPrefer not to say
Jun 2026

Summary

Google ML Engineer system design round, focused entirely on large-scale video deduplication. Heavy on architecture and trade-offs, less on ML theory than I expected. Left feeling like I talked too much about embeddings and not enough about the operational side.

Questions Asked (1)

Q1

Design a system that can detect and remove near-duplicate short videos at massive scale, handling millions of uploads per day.

System DesignTechnical Trade-offsAlgorithms & Data Structures
Author's notes

I jumped straight to ViT embeddings and vector search, which felt right, but I spent way too long on the feature extraction part and barely touched ingestion throughput or queue design.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying requirements and scale, then propose a multi-stage pipeline: efficient candidate generation via hashing or embeddings, followed by precise verification. Discuss trade-offs between recall, precision, latency, and cost, and how to handle scale with distributed systems.

Pro tip: Emphasize that near-duplicate detection is not just about exact matches; it's about semantic similarity. Mention that you would use a combination of perceptual hashing for speed and deep learning embeddings for robustness, and always consider the feedback loop from user reports.

1. Clarify Requirements and Scale

Ask about definition of near-duplicate, acceptable false positive/negative rates, latency requirements, and scale (millions/day). Confirm whether detection is real-time or batch.

2. Design Candidate Generation

Propose efficient methods to reduce the search space, such as perceptual hashing (pHash, dHash) or locality-sensitive hashing (LSH) on video fingerprints. Discuss how to handle short videos (e.g., keyframe extraction).

3. Verification and Similarity Scoring

For candidate pairs, use more expensive methods like deep learning embeddings (e.g., from a video model) and compute similarity (cosine, Euclidean). Set thresholds based on precision/recall trade-offs.

4. System Architecture and Scalability

Outline a distributed pipeline: ingestion, feature extraction, indexing (e.g., FAISS, Annoy), and decision service. Discuss partitioning, sharding, and use of cloud services (e.g., Bigtable, Dataflow).

5. Handling Removal and Feedback

Describe how to remove duplicates (e.g., keep original, remove later uploads) and incorporate user feedback and appeals. Discuss monitoring and iterative improvement.

Key Points to Mention

  • Perceptual hashing (pHash, dHash) for fast approximate matching
  • Locality-Sensitive Hashing (LSH) for efficient nearest neighbor search
  • Deep learning embeddings (e.g., from 3D CNNs or transformers) for semantic similarity
  • Trade-offs between precision and recall, and how to tune thresholds
  • Distributed systems for scale: sharding, batch vs. stream processing
  • Feedback loops and continuous learning from user reports

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.