CV Parser and AI Matching

We need to parse Resume/CV and extract the document into structured information in JSON or XML format — suitable for storage, reporting and manipulation. Input:This parser will be used to parse thousands of UNSTRUCTURED resumes in html, word (doc, docx), rtf, text, rtf and pdf formats. Resumes will be in Portuguese language to start with but the parser should be capable of parsing resumes in other languages. Output: JSON or XML format files of the resume when all the words from resume are parsed and structured correctly. Parser should also output the report and metric on what worked and what did not. This report should help in further tuning the parser. Accuracy: We areexpecting a high degree of accuracy. Ideally the parser should be able to reach maintain 95% accuracy. We are expecting minimum of 90% accuracy for each resume. Method: We are not looking at keyword based parser. We are looking for a parser that can be trained and improve its accuracy over a period. Also, the fields parsed should be configurable and we should be able to extend the parser in a fashion that we can add more fields to be parsed with little effort. Parser should be able to do Fuzzy look up Resume Parser should be written such that in can eventually be invoked using a web based application developed in .NET c#. In order to identify information from CVs and profiles, the extraction engine should learn to create “rules” within its machine-learning algorithms by analysing the data. Fields that need to be parsed – Not all the resumes will have all the fields – Also the parsing should not limit to just these fields. 1. First Name 2. Middle Name 3. Last Name 4. Email Address 5. Mobile Number 6. Date of Birth 7. Picture/Image 8. Social media profile links 9. Gender 10. Nationality 11. Address 12. City 13. State 14. Country 15. Zip Code 16. Summary/Objective 17. Achievements 18. Current Location 19. Skills(top to bottom) 20. Current Company Name 21. Current Company Designation 22. Current Company Start Date 23. Current Company End Date 24. Current Company Roles & Responsibilities 25. Previous Company Name 26. Previous Company Designation 27. Previous Company Start Date 28. Previous Company End Date 29. Previous Company Roles & Responsibilities 30. Education 31. Year of passing 32. Board/university 33. Certification 34. Marital Status 35. Language Known 36. Interest/Hobbies 37. References 38. Soft Skills 39. Training courses Output of the parser should be an xml or JSON tagged file, one xml file for each parsed resume, output file name to be the same as the input file name with extension. All the parsed fields will be used to upload into a SQL database. Parser is also required to do the database insertion as part of the parsing process. It should also be possible to obtain the best candidate for an opportunity through an intelligent algorithm to be validated in a future meeting We already have a database with about 40,000 resumes and the system should be able to learn the resumes pattern to improve results We will supply a sample set of resumes, as many as you need to be successful.