Ebonite is an open-source framework for managing machine learning models. It is developed by a team at https://zyfra.com/
With ebonite, you can create Docker containers from your models with just a couple lines of code! But also ebonite is very flexible and fully customizable.
This project started in late 2018 as an in-house tool for fast deployment at Zyfra AI. Later Mikhail Sveshnikov gave a talk at DataFest6 about it (here) and we decided to opensource this project. The first public release was ready in November 2019, and since then everybody can participate in development.
As of summer 2020, the project is still actively developed by a team at Zyfra AI. However the 1.0.0 release is still in the distant future, so all of the APIs must be considered not stable/experimental. But because of it, it's the perfect time to make contributions that will have a lot of impact on the project's future.
Mikhail Sveshnikov is a technical lead of the project, so in case you want to participate in the development, hit him up in slack / telegram / github.
Kirill Vlasov is a product owner, contact him for business queries/applications.
