Company
CFT
8 cases from CFT (General, Fintech, Healthtech), each digested from a public talk or engineering post and linked to the original. 3 report results in production.
Tasks: Match entities, Get experiment results faster, Avoid false wins in A/B tests, Catch fraud and abuse, Moderate content at scale and 1 more. Methods: Data augmentation, Speech recognition, HNSW, Contrastive learning, Embedding fine-tuning, k-nearest neighbors and 26 more. Tools: BERT, Llama, Elasticsearch, multilingual-E5, Hugging Face, Mistral and 12 more.
| Shazam for Full Names: Finding a Match When a Name Can Be Written 20 Different WaysCF | CF | fintech | 2024-05 |
| LLM Practice Without ChristofariCF | CF | general | 2024-05 |
| So A/B Tests Are Just OLS Regression?CF | CF | general | 2024-05 |
| Learning to Understand Dialogs Based on Real DataCF | CF | fintech | 2023-10 |
| Please Don't Shout. ML vs Operators: 1-0CF | CF | general | 2023-05 |
| Solving the Face Spoofing Problem in PhotographsCF | CF | fintech | 2023-05 |
| Improving speech recognition without labels, sign-up or SMSCF | CF | general | 2022-10 |
| Are Vaccines Effective? Looking at Publicly Available DataCF | CF | healthtech | 2022-06 |