Published offering
Appen describes speech corpora, image and video annotation, agent trajectories and model evaluation. Dataset availability and acceptance criteria should be agreed for the specific engagement.
Questions for a project brief
- What is the source of each dataset, and what evidence supports the proposed use?
- Which languages, domains and edge cases are included or missing?
- How are annotators selected, instructions tested and disagreements reviewed?
- Can you inspect a representative sample and agree acceptance checks before committing?
- What do the delivery terms cover: formats, updates, exclusivity and model-training use?
Provider descriptions establish what is offered, not whether a particular dataset is suitable for your workload. Dataset-specific terms and quality still need review.