Tool
UMAP
10 cases from 8 companies use UMAP in their stack. 1 was checked in an online A/B test; 1 report results in production. Tasks: Learn user and item embeddings, Recommend to new users and new items, Recommend from what the user just did, Analyze user behavior and 3 more. Methods: UMAP, HDBSCAN, Gradient boosting, Matrix factorization, Text clustering and 30 more. Companies: MegaFon, VTB, Spotify, Whatnot and 4 more.
Unsupervised Topic Discovery with Text Embeddings![]() | ![]() | fintech | 2026-10 |
A Cluster-Based Approach to CTR Forecasting for Millions of Advertised Products![]() | ![]() | marketplace | 2026-05 |
| LLMs and Other NLP Models in the Service of CRM and Customer Journey Improvement | banking | 2025-05 | |
| Learning From Structure: Discord’s Entity-Relationship Embeddings | social | 2024-09 | |
| Recursive Embedding and Clustering | media-streaming | 2023-12 | |
| AML Reimagined: LaundroGraph Exploits Graph Structure to Assist Anti-Money Laundering Activities | banking | 2023-09 | |
Clustering subscriber messages![]() | ![]() | telecom | 2022-06 |
MegaFon RecSys Workshop, part 2: a two-level recommender (LightFM + CatBoost) and DSSM models![]() | ![]() | media-streaming | 2022-06 |
MegaFon RecSys Workshop, part 1![]() | ![]() | media-streaming | 2022-06 |
| Going from 0 to 1 modeling User Preferences for Personalized Recommendations | marketplace | 2022-03 |


