Data work runs on shared languages, from gesture and speech to charts and code. Through them, people and machines understand one another and work together. What we can ask of data and do with it together reaches only as far as these languages do. My work explores them: how we use them, how we can improve them, and what new ones might make possible.
For example, I build meeting notes where every “this” and “here” replays the deictic gesture behind it (Deixis), presentation tools that turn sketches and spoken questions into live analysis (SlideSAVR), and representations that make AI-generated code reviewable when human attention is scarce (.reviewme). I also study the languages themselves: how to define community-shaped chart types whose names cover varied designs (chart type boundaries), and how a domain-specific language’s design invites the mistakes AI makes in it (DSLs for LLMs).
Through my work, I hope people keep an irreplaceable place in data work and beyond, where labor can become “not only a means of life but life’s prime want.”