Appen

Appen provides the human-produced data that AI models learn from. Its work has moved from speech and search quality assessment to data for modern language and multimodal models.

What Appen offers

• AI training data — datasets annotated by people to teach models about language, images and speech.
• Data annotation — labelling of text, images and other content at scale.
• Model evaluation — expert review of how AI systems perform.
• Off-the-shelf datasets — ready-made data collections alongside bespoke enterprise projects.
• Workforce programme — a global pool of contributors who can apply to join.

Who it works with

Appen supplies teams building AI products, such as search engines, voice recognition and social platforms. Its contributors work remotely.

Key facts

• Experience: about 30 years in data work
• Reach: operations in more than 130 countries
• Languages: over 180 languages and dialects
• Workforce: more than one million flexible workers, according to earlier company text