Greater Parsing Technology with GPT-3

CVandMe is an entirely new CV/resume parser built with GPT-3.

It provides the highest accuracy levels to date and strongly improves modern applicant tracking systems (ATS) and job marketplaces.

Test for yourself.





Power new features by removing parsing errors

Current parsers do not reliably detect all candidate experiences or wrongly introduce non-existing ones. This means that your current features do not work effectively. Also, you might not even have launched features that can increase your software's potential in automating recruitment processes or improving the candidate experience. Our parser enables:
  • More job applications: with precise pre-population of application forms job candidates need to enter less data manually leading to more completed applications
  • Faster hiring: Detect great candidates automatically and invite them to an interview minutes after their application
  • Find suitable candidates from your CV database of internal and external candidates
  • Re-skilling / up-skilling: Understand and map capabilities of current employees
  • Get analytics on your candidates and remove bias




Leveraging deep learning advances of 2019-2021

CVandMe has built its parser from scratch using
  • GPT-3 language models finetuned on proprietary data
  • Deep image segmentation and classification models
GPT-3 is one of the best deep learning models to understand text. It was developed by OpenAI and has 175 billion parameters. CVandMe has been part of the few beta testers since its launch in 2020. We also employ deep learning models to detect visual features of single words in a CV and to segment a CV into logical areas.

CVandMe is superior for parsing
  • CVs in image format (jpg, png) or in pdf-format with text embedded as images
  • CVs with 2 or 3 columns
  • Distinguishing major skills (developed frontend in JS) from minor ones (reported on JavaScript project)
 
 

Company

CVandMe is a brand of Candidate & Company GmbH, headquartered in Berlin (Germany). It was founded in 2020 by machine learning experts and serial entrepreneurs. We use the latest technology breakthroughs in natural language processing to significantly improve organizational processes.

Simon Krueger

Simon Krüger  

Co-founder

Serial entrepreneur,
co-founded Doctolib Germany

Trian Xylouris

Dr. Trian Xylouris  

Co-founder

Ex-McKinsey consultant,
math PhD & state-of-the art Linnik's theorem


Join us!

We are looking for ambitious talent in machine learning research and engineering. If you have expertise in document parsing, GPT-3 or image segmentation, we look forward to hearing from you at [email protected]