← Meta Interview Insights

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

Senior
Jun 2026

Summary

ML system design round at Meta for an engineer role. One meaty open-ended question about building a weapons ad detection system from scratch, no warmup, just straight into it.

Questions Asked (1)

Q1

Design a complete machine learning system to automatically detect advertisements that are selling weapons.

System DesignTechnical Trade-offsProduct Sense & Ideation
Author's notes

This is the kind of question where you can spiral fast if you don't anchor early.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the scope: define what constitutes an advertisement selling weapons, including policy nuances (e.g., legal vs. illegal weapons, context like toys or video games). Then outline an end-to-end ML system covering data collection, labeling, feature engineering, model selection, deployment, and monitoring, while addressing trade-offs like precision vs. recall and scalability.

Pro tip: Emphasize a human-in-the-loop approach and multi-modal signals (text, image, video) to improve accuracy and adapt to evolving adversarial tactics, showing you understand real-world content moderation challenges at scale.

1. Clarify Requirements and Scope

Ask questions to define what 'advertisements selling weapons' means: types of weapons (firearms, explosives, etc.), legal vs. illegal, and platforms (feed, marketplace, ads). Establish success metrics like precision, recall, and latency.

2. Data Collection and Labeling

Identify data sources: ad text, images, videos, landing pages, and user reports. Design a labeling schema with policy experts and use techniques like active learning and weak supervision to scale labeling.

3. Feature Engineering and Model Selection

Extract multi-modal features: text (keywords, embeddings), image (object detection, OCR), and metadata. Choose models like fine-tuned transformers for text, CNNs for images, and ensemble or multi-modal fusion for combined signals.

4. Training, Evaluation, and Trade-offs

Train models with class imbalance handling (e.g., focal loss). Evaluate using precision-recall curves, and set thresholds based on business cost of false positives vs. false negatives. Consider adversarial robustness and bias mitigation.

5. Deployment, Monitoring, and Iteration

Deploy as a real-time service with a human review queue for borderline cases. Monitor performance drift, adversarial attacks, and feedback loops. Continuously retrain with new data and update policies.

Key Points to Mention

  • Multi-modal approach combining text, image, and video analysis for higher accuracy.
  • Handling class imbalance and setting appropriate precision/recall trade-offs based on policy.
  • Human-in-the-loop system for reviewing edge cases and providing feedback.
  • Scalability and low-latency inference for real-time ad detection at Meta's scale.
  • Adversarial robustness: detecting obfuscated language, coded terms, and evasive imagery.
  • Ethical and legal considerations: avoiding over-censorship, bias, and ensuring compliance with laws.

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