Open Source Machine Learning
Welcome to Seldon, your full-stack open-source machine learning solution for building recommendation engines, creating intelligent apps and services, and gaining actionable insights.
As a developer or data scientist, you trust a platform that scales horizontally in production in the most demanding enterprise environments. You need total control to customise and operationalize your models faster, with no lock-in and the freedom to deploy securely on-premise or in the cloud.
At Seldon, one of our core principles is to give you the best tools for the job, and that’s why our microservices architecture makes it easy to extend the platform with open-source libraries or even closed APIs. We have released microservices to integrate Vowpal Wabbit and XGBoost, Keras and Gensim.
Our team and community are passionate supporters of open source technologies, and that’s why Seldon is powered by the cutting edge of open source data science projects such as Spark, Kafka, InfluxDB and Zookeeper.
Our mission is to build a fully integrated open-source ecosystem by providing an enterprise-grade machine learning solution that is totally customizable to your requirements and helps you to solve real-world problems faster.
Customisable to support your use cases and internal systems
Deploy behind your firewall on your internal servers or private cloud
Performance testing and optimisation tools
Move faster with Seldon on your data science team
- Extend your data science team with enterprise support.
- Flexible deployment options with clear migration path.
- Reduce development time and costs: machine learning algorithms from the open-source community, Seldon and your in-house data science team.
- Move faster with industry models, customizable to support your use cases and internal systems.
- Performance testing and real-time optimisation with multi-armed bandit.
Explore Seldon’s infrastructure
Seldon’s highly scalable platform is dockerized to simplify deployment to your private cloud or dedicated server infrastructure all running within a Kubernetes cluster. Our open predictive platform offers two main features, content recommendations and predictions.