Company
Yelp
10 cases from Yelp (Marketplace, Adtech, Search engine), each digested from a public talk or engineering post and linked to the original. 1 was checked in an online A/B test; 2 report results in production.
Tasks: Rank search results, Build a data platform, Catch fraud and abuse, Moderate content at scale, Build an assistant over internal docs and 8 more. Methods: Supervised fine-tuning, Threshold selection, Embedding similarity matching, RAG, Few-shot prompting, Monte Carlo / what-if simulation and 27 more. Tools: MLflow, XGBoost, OpenAI API, Hugging Face, Prometheus, LangChain and 15 more.
Tech blog: engineeringblog.yelp.com
| ML based ranking using Nrtsearch | search-engine | 2026-09 | |
| Building Biz Ask Anything: From Prototype to Product | marketplace | 2026-03 | |
| How Yelp Built a Back-Testing Engine for Safer, Smarter Ad Budget Allocation | adtech | 2026-02 | |
| Search Query Understanding with LLMs: From Ideation to Production | marketplace | 2025-02 | |
| dbt Generic Tests in Sessions Validation at Yelp | general | 2024-08 | |
| Moderating Inappropriate Video Content at Yelp | marketplace | 2024-03 | |
| Yelp’s AI pipeline for inappropriate language detection in reviews | marketplace | 2024-03 | |
| Yelp Content As Embeddings | marketplace | 2023-04 | |
| Lessons from A/B Testing on Bandit Subjects | adtech | 2022-12 | |
| Beyond Matrix Factorization: Using hybrid features for user-business recommendations | marketplace | 2022-04 |