Textract: Extract text from a large variety of file formats

I'm assuming that this was the intended link:

https://aws.amazon.com/textract/

This one’s interesting, because it seems to support more formats than Apache Tika and even includes speech recognition and OCR, all conveniently rolled into one package.