Lilac

Lilac is an open-source tool built for data and AI practitioners who know that better models start with better data. If you've ever spent hours trying to find specific examples in a massive dataset, identify duplicate entries, or root out sensitive PII before training an LLM, Lilac was designed with you in mind. It offers semantic and keyword search, field editing and comparison, PII detection, duplicate identification, language detection, custom signal integration, and fuzzy-concept search with iterative refinement. It's the data quality toolkit that makes preparing LLM training data faster, more reliable, and less painful.