Hello Bharath,

Your concerns are justified. Rule based parsers do have limitations no doubt. In our case we had requested the candidates to send their resumes by prefixing their names. for e.g ‘Bharath_resume’. So our jobs got easier. If you noticed, we are extracting names from the resume file name itself. we are not parsing the contents of the resume to get candidate name.

Alternatively, one could also parse the contents of the resume and extract entities like names, titles and organisation details. This would require you to custom train your own neural network model.