In order to make Machine Learning suitable for the masses and accessible to a wide range of users and applications, as well as to bring real added value, its development must be cost-effective.
An ecosystem that wants to achieve this must remove the currently most important limitations in AI development:
We at MLReef have defined these points as our goals.
MLReef is an MLOps platform for efficient, collaborative and replicable work on Machine Learning (ML) projects. It is globally the first to set its core element on community collaboration to enable instant reuse of any ML element previously published on MLReef. This drastically shortens development time and increases the model quality through fast and structured iterations.
This unique selling proposition means that user generated public data, AI models, data visualization and data processing can be used immediately by everyone in a single environment. This network effect accelerates the innovative power of our target groups enormously, since ML projects do not always have to be developed from scratch but can build on existing projects very easily and quickly.
MLreef is designed to work for a broad range of customer segments, from academia, individual data scientists, small teams (2 – 3 individuals) and enterprises.
Data gathered by ESA´s Earth Observatory (EO) is an ideal base for many scientific and commercial ML applications. In this context, MLReef provides a short bridge to a holistic ML environment to create downstream research and commercial application. The aim at ESA BIC is to define and develop technical data integrations with leading EO based data providers (EODC, TU Vienna, DIAS).