🔧 Installation • ⛽ Servers • 🤖 Models • 🔗 Pipelines • 🧑🔬 Experiments • 📊 Performance Tuning
## 🧩 Features * **Pipelines**: Deploy composable AI applications, leveraging Kafka for realtime data streaming between components * **Autoscaling** for models and application components based on native or custom logic * **Multi-Model Serving**: Save infrastructure costs by consolidating multiple models on shared inference servers * **Overcommit**: Deploy more models than available memory allows, saving infrastructure costs for unused models * **Experiments**: Route data between candidate models or pipelines, with support for A/B tests and shadow deployments * **Custom Components**: Implement custom logic, drift & outlier detection, LLMs and more through plug-and-play integrate with the rest of Seldon's ecosytem of ML/AI products! ## 🔬 Research These features are influenced by our position paper on the next generation of ML model serving frameworks: 👉 [Desiderata for next generation of ML model serving](http://arxiv.org/abs/2210.14665) ## 📜 License Seldon is distributed under the terms of the The Business Source License. A complete version of the license is available in the [LICENSE file](LICENSE) in this repository. Any contribution made to this project will be licensed under the Business Source License.