I would use Google Vision Document Text Detection over Tesseract OCR. It’s very accurate and is run in Google Cloud so no need to bog down your machine. https://cloud.google.com/vision/docs/pdf