Argilla

Argilla is an open-source data curation platform for large language models (LLMs). Argilla aims to help users build language models faster through data curation with both human and machine feedback, providing support for steps in the MLOps cycle, including data labeling and model monitoring. Argilla is compatible with major NLP libraries, including Hugging Face transformers, spaCy, Stanford Stanza, and Flair. Users can combine their preferred libraries without having to implement a specific interface. Argilla allows users to combine hand-labeling with active learning, bulk-labeling, zero-shot models, and weak supervision in novel data annotation workflows.

Argilla is built on five core components:

1. Python SDK—To interact with the Argilla Server and the Argilla UI. It provides an API to manage the data, configuration, and annotation workflows.
2. FastAPI server—A Python FastAPI server that manages the data, pre-processing it and storing it in the vector database. It also stores application information in the relational database. It provides a REST API to interact with the data from the Python SDK and the Argilla UI. It also provides a web interface to visualize the data.
3. Relational database—A relational database to store the metadata of the records and the annotations. SQLite is used as the default built-in option and is deployed separately with the Argilla Server. A separate PostgreSQL can also be used.
4. Vector database—A vector database that stores the records data and performs scalable vector similarity searches and basic document searches. Argilla supports ElasticSearch and AWS OpenSearch, and they can be deployed as separate Docker images.
5. Vue.js UI—A web application to visualize and annotate data, users, and teams. Built with Vue.js, it is directly deployed alongside the Argilla Server within our Argilla Docker image.

Argilla is a Spanish company that was founded in 2017 by Daniel Vila Suero (CEO) and Francisco Aranda. The company was originally called Recognai, and its first product, launched in 2021, was named Rubrix. In October 2022, Recognai rebranded its name and platform to Argilla. The new name comes from the word "clay" in Latin and the idea of modeling clay and shaping models. Argilla has locations in Madrid and Carpresa, Valencia.

On January 25, 2023, Argilla announced $1.6M in funding from Zetta Venture Partners (Kaggle, Domino Data Lab, Weaviate) and Caixa Capital Risc (Vilnyx, Codee).