Loading
Raptor (RaptorML) is an Apache-2.0 open-source project. It exports Python feature and model code as production artifacts and deploys them to Kubernetes, SageMaker or Docker. The runtime handles data connections, caching, scaling and monitoring, and it aims to remove training/serving skew without a separate feature store. It was created by Almog Baku, who was developing it by 2022. The GitHub repo has about 150–160 stars, and the latest Go module release (v0.3.3) was published in April 2024. I found no public funding, revenue, headcount or commercial offering.