Data built for real-world AI applications

Pick the kind of AI you are building. Each page covers the data it needs, how we collect it and who uses it.

Build data for AI assistants, chatbots and voice interfaces.

Speech data for ASR, voice interfaces and language understanding.

Human-generated and evaluated data for modern generative AI systems.

Language data for training, evaluation and language understanding.

Image and video data for visual AI applications.

Structured document data for extraction and intelligent document processing.

Data connecting text, speech, image and video for multimodal systems.

Human evaluation workflows for measuring AI system outputs.

Building something that isn't listed?

Describe the model and the failure you are trying to fix. We will tell you what data would help.