...

Workee

A mobile first service for connecting blue collars with hiring employers

...

Challenge

The mail goal of CactusSoft’s developers was to create their own recruiting app for quickly fulfilling critical positions. The search results had to be based on the location of candidates. The main challenge was to develop a full-text search in Danish and English, which would allow users to find both employees and vacancies not just for a particular query, but also taking into account synonyms and cognate words.

Solution

The CactusSoft team developed prototypes of the required search system using two solutions – MongoDB and ElasticSearch. During test cases, the ElasticSearch system showed better results, so it was integrated into the application.

In addition, developers created a trainable neural network, which analyzes the lists of companies and candidates and proposes an optimal “candidate–vacancy” pair. It helps an employer who has published a new position get a list of top candidates who meet their requirements.

In addition, developers created a trainable neural network, which analyzes the lists of companies and candidates and proposes an optimal “candidate–vacancy” pair. It helps an employer who has published a new position get a list of top candidates who meet their requirements.