Company
Wildberries
9 cases from Wildberries (Marketplace, E-commerce), each digested from a public talk or engineering post and linked to the original. 2 report results in production.
Tasks: Catch fraud and abuse, Moderate content at scale, Build an assistant over internal docs, Evaluate models, Build a data platform and 4 more. Methods: Feature store, Supervised fine-tuning, Contrastive learning, Embedding fine-tuning, Vector / semantic search, Vision-language model and 47 more. Tools: Amazon S3, ViT, Mistral, multilingual-E5, CatBoost, Qwen and 36 more.
High-Speed Multimodal Moderation: LLMs and VLMs in Real Production![]() | ![]() | marketplace | 2025-05 |
An Effective Approach to Quality Control of Anti-Fraud Models![]() | ![]() | marketplace | 2025-05 |
Increment vs Experiment: How to Reduce Uncertainty When Developing ML Models![]() | ![]() | e-commerce | 2025-05 |
An LLM Reads Marketplace Reviews for You![]() | ![]() | marketplace | 2024-05 |
Optimizing Marketplace Operations with Language Models![]() | ![]() | marketplace | 2024-05 |
How We Fought Spam in Reviews![]() | ![]() | marketplace | 2024-05 |
RetrievalRecsys: An Architecture Modification to Account for Text in Personal Recommender Systems![]() | ![]() | marketplace | 2024-05 |
Advanced RAG Pipelines![]() | ![]() | marketplace | 2024-05 |
ML Module Architecture for Automating Technical Support for Pickup Point (PVZ) Staff![]() | ![]() | marketplace | 2023-10 |
