# super.AI vs Hyperscalers

What’s the difference between Super.ai and Hyperscaler Intelligent Document Processing (IDP) solutions? Let’s take a look at how the features of both approaches compare.

## How super.AI compares to Hyperscalers

Includes: Google Document AI, Microsoft Syntex, AWS Textract, IBM Watson Discovery

### Technical

| Metric                                | super.AI | Hyperscalers |
|---------------------------------------|----------|--------------|
| Accuracy range                        | high     | low          |
| Automation level                      | high     | medium       |
| Handling changing formats              | excellent | partial      |
| Supported format types                | high     | low          |
| Custom fields                         | high     | medium       |
| Trainable                             | high     | high         |
| Scalability                           | high     | low          |
| Flexibility                           | high     | medium       |
| Expandability                         | high     | high         |
| Auditability                          | high     | low          |

### Business

| Metric                                | super.AI | Hyperscalers |
|---------------------------------------|----------|--------------|
| Ease of use                           | high     | low          |
| Automation ROI                        | high     | medium       |
| Adherence                             | high     | high         |
| Time to value                         | fast     | slow         |
| Time to value (ROI)                  | high     | low          |
| Rollout requirements (lower is better)| low      | low          |
| Security & compliance                 | high     | low          |
| Future-proofing                       | high     | low          |
| Processing cost                       | competitive | competitive|

## Intelligent Document Processing (IDP)

Automate information extraction from any business document with guaranteed quality, backed by human review and Data Processing Crowd.