Thank you for your request😊
We will consider it within an hour during our business hours 🏃and then contact you with a stack of cool solutions!💪🏻
JobSite Service
a web-based recruiting service with a unique search algorithm
We had such
development tasks:
development tasks:
- Develop an algorithm for selecting applicants for the criteria of vacancies
- Create a mechanism for the intellectual analysis of resumes taking into account the preferences of the employer
- Develop an algorithm for processing and outputting statistical data
- Develop a chatbot for the automated polling of potential candidates
JobSite technologies
Final Result
Jobsite search algorithm
Jobsite is a web-based recruiting service with a unique search algorithm.
For the successful implementation of the project, we developed our own algorithms
based on graphs, where the top of the graph consists of facts about the applicant
and the employer and their interests, and the “edges” act as the connection between
them. Then, according to the specified criteria of the vacancy, our system selects
precisely those candidates who are most accurately correlated according to the necessary
parameters.
The next task was to create a mechanism for the intellectual analysis of resumes. Our
system has the ability to train the data entering it, with given initial parameters.
Learning is carried out using various methods and algorithms of machine learning, in
particular the methods of so-called “reinforcement learning”, “learning without a teacher”,
algorithms for automatic clustering of words based on vectorization (glove, word2vec, fasttext),
etc.
Our client was interested in an efficient mechanism for processing statistics. To successfully
resolve this problem, we combined object-relational and network database types, and also actively
used partitioning and intelligent caching. We use dozens of servers to calculate the relationship
between candidates and employers. For each work, we display comprehensive statistics about the
data collected, but this doesn't affect the speed and performance of the resource.
The next task was to develop a chatbot for interviewing the applicant. We've developed
a humanoid chatbot, which offers the applicant to participate in a survey about hiring.
The structure and questions of texts are dynamically changed by portal administrators.
As the candidate is interviewed, our system comprehensively analyzes messages, determines
the tonality of the text and sets the compatibility rating for this vacancy.
You give us an idea?
we bring the result
we bring the result
