Argilla

Argilla is a collaboration tool for AI engineers and domain experts who need to build high-quality datasets for AI projects. It is used for collecting human feedback and labeling data for NLP, LLM, and multimodal workflows, with a focus on continuous evaluation and model improvement.

It lets teams interact with data through filters, AI feedback suggestions, and semantic search. The programmatic approach supports text classification, NER, RAG, preference tuning, and other text labeling and annotation tasks, so teams can iterate on the right data and models.

Argilla is open source and now in maintenance mode, with the maintainers continuing to publish bug fixes and patches against a mature, stable codebase. You can deploy it on Hugging Face Spaces or run it from source on your own infrastructure, keeping ownership of your data and models.