- The increasing need for larger volumes of training data to support their programs — especially to tune the models they’ve already deployed
- Quality and speed requirements — needing properly structured data, fast
- The need for broad datasets that cover diverse real-world scenarios — especially concerning translation data for under-resourced languages, chatbot training to better route support requests, and more
- A heightened need for data security and privacy — especially when dealing with PII, payment data, and medical imaging data
![Woman standing at Appen booth](https://s41256.pcdn.co/wp-content/uploads/2019/04/D3v1ht6UwAEACmX-750x450-1.jpg)