Overview

Our client is one of the leading resource augmentation service providers based out of US. They offer manpower services for different positions and across different industries. They process several resumes in a day with docx, doc and rtf formats using existing system and the output will be delivered in HRXML format. 

Client:
ProRank
Industry:
Human Resources
Services:
Resource Augmentation
Website:
prorank.ai
Technology:
Django,PostgreSQL,Flask

Challenge

Dependency:

The client was fully reliable on the existing system which parses resumes only in three different formats.

Non Customizable:

The HRXML output from the existing system could not be customized as per the client’s requirement.

Time Consuming:

The parsing time taken by the existing system for a resume remained as an issue, as the volume of resumes to be parsed in a day was higher.

Solution

We at Agira, carefully analyzed the existing system and client’s requirement and decided to build a new application,

  • The new application was architectured into two components,

Web based UI

Parsing Engine

  • The UI was designed with the help of ‘Django’.
  • Users could easily upload the resumes from UI and view the different sections of resumes.
  • Advanced filtering features were built to let users to customize the candidate parameters for parsing the resumes.
  • Parsing engine was built using Regular Expression (RE) module in Python.
  • This enabled users to choose resumes in any of these formats[.docx, .doc, .rtf, .txt, .pdf, .html]
  • With the help of ‘Postman’, the output could be generated in customizable ‘HRXML’ format and ‘JSON’ format
  • ‘Postgresql’ database was used to handle the parsed data.

Benefits

The client was delighted with the new application built. Some of the key areas that helped their business,

The entire application was built in the time span of 3 months and thus helped the client to migrate to the new system

The new application benchmarked an accuracy rate of 70% compared to existing system, which is higher than what is expected by the client

Performance time is highly efficient and lies between 5-7seconds for handling both 20 as well as 12,000+ resumes at the same time.

The new application has value added feature to compare the parsed resumes with other systems.