We believe in the power of collaborative, systematic and reproducible R&D to solve the World's most challenging problems!

Check out our RESCUE-HPC workshop at SC18!
See CK use-cases from General Motors, IBM, Amazon, Arm, RPi, ACM (DNN SW/HW co-design), Supercomputing, CGO and dividiti.
Collective Knowledge (CK) is an open-source framework to speed up collaborative and reproducible research with reusable, customizable, portable and non-virtualized (!) components (code, data and models). Trusted by a growing number of academic and industrial partners, CK unifies and dramatically accelerates complex experiments such as benchmarking, co-design and optimization of the whole SW/HW stack for AI/ML from cloud to edge. Give it a try!

CK is a community-driven, decentralized, small and cross-platform framework to help users organize their local code, data and scripts; convert them into portable, customizable and reusable components with a Python API, JSON meta description and an integrated package manager; quickly prototype research workflows on Linux, Windows, MacOS and Android; automate and crowdsource complex experiments; generate interactive papers; support reproducible tournaments such as ACM ReQuEST and accelerate open science!

Get started