This shares similarities with Kubeflow/Airflow but with a better UX. BYO infra.
Quoted:
Shopify's Search & Discovery team built Tangle (originally Cloud Pipeline Studio, a tool I built before joining Shopify) to accelerate ML experimentation without sacrificing reliability.
Tangle is proving to be useful for teams all across Shopify. It works at our scale. Data scientists iterate faster. Engineers maintain less custom pipeline code. Team members share pipelines and runs. Infrastructure costs decrease through global caching.
Shopify is committed to making the web as good as it can be through open-source contributions. We know the broader ML community faces identical challenges to us, so open-sourcing Tangle was a no-brainer. This way, we can extend Tangle’s impact and invite collaboration from researchers and practitioners worldwide.
vincentzed•2mo ago
Quoted: Shopify's Search & Discovery team built Tangle (originally Cloud Pipeline Studio, a tool I built before joining Shopify) to accelerate ML experimentation without sacrificing reliability.
Tangle is proving to be useful for teams all across Shopify. It works at our scale. Data scientists iterate faster. Engineers maintain less custom pipeline code. Team members share pipelines and runs. Infrastructure costs decrease through global caching.
Shopify is committed to making the web as good as it can be through open-source contributions. We know the broader ML community faces identical challenges to us, so open-sourcing Tangle was a no-brainer. This way, we can extend Tangle’s impact and invite collaboration from researchers and practitioners worldwide.