Faster, cheaper, more trustworthy data analysis for humans and AI agents. A Python library that turns data-science scripts into a cached, reproducible, lineage-tracked pipeline — reruns only what a parameter, data, or code change affects. No server, no account.
