is an open platform developed by Grigori Fursin to support the Collective Knowledge project (CK) and reproducibility initiatives started in 2015. The long-term goal is to enable efficient and self-optimizing software and hardware for emerging AI, ML and quantum workloads. We want to achieve that by providing a unified way to package and share artifacts and ML models along with papers. The idea is to help researchers and practitioners reproduce results and participate in collaborative ML/SW/HW benchmarking and optimization using "live" scoreboards.

This is still a prototype and we are working to provide a GUI to assemble new CK solutions and pack your ML models and artifacts in a user-friendly way.