Skip to main content
All CollectionsProduct & Features
How does the Hirize Parser Work?
How does the Hirize Parser Work?
Updated over a year ago

Using a combination of Natural Language Processing (NLP), Optical Character Recognition (OCR), and Computer Vision, our parsing engine seamlessly processes a wide range of document formats, including DOC, DOCX, PDF, RTF, TXT, HTML, and image-based formats like JPEG or PNG.

Here's how it works:

  • Data Input: Simply provide your document in any supported format, and our parsing engine will handle the rest.

  • Preprocessing & OCR: If your document is in a non-text format or an image, our OCR technology will kick in and convert it into machine-readable text. At the same time, the engine carries out necessary preprocessing tasks like text normalization.

  • Tokenization: The engine breaks down the text into individual words or "tokens," setting the stage for the next step.

  • Parsing & Computer Vision: Using sophisticated machine learning algorithms and predefined rules, the engine categorizes each token. It recognizes names, job titles, skills, company names, dates, and more. Computer Vision techniques come into play for complex and differently structured documents.

  • Post-Processing: The categorized data is structured into a standard format, allowing for easy interpretation and further processing. The engine generates a structured JSON document with fields for name, contact info, work history, skills, education, and more.

  • Output: The parsed data is then provided as output, ready to be utilized by other software or stored in a database.

Our parsing engine is constantly evolving, learning from its past performance, and improving its accuracy. With Hirize, you can trust in the highest levels of accuracy and efficiency in resume and job description parsing.

Did this answer your question?