Using programming language techniques to make it easier,
safer, and faster to ingest
untrusted or ill-formed data.
The Fisher Lab uses programming language techniques such as domain-specific languages, program synthesis, and formal methods to make it easier, safer, and faster to ingest untrusted or ill-formed data from the outside world into a program and then to efficiently compute with that data once it is in-memory. This kind of data wrangling is an oft-overlooked early step in many data science pipelines.