Proscom AI Animator Service for HR Specialists

The main articles are:

• Machine Learning
• Data mining Data mining
• Big Data
• Speech Recognition (Technology, Market)
• Speech technology: On the path from recognition to understanding

2024

Name change

Until June 2024, it was called Proscom Hrobot AI service for HR specialists.

Announcement of a solution for optimizing the work of HR specialists

On January 23, 2024, IT company Proscom announced the creation of a solution to optimize the work of HR specialists. The Nanimator service will optimize the initial processing of applicants' resumes and simplify the selection of candidates. This ML project (Machine Learning) was originally implemented for the needs of the company, but after 5 months it showed effectiveness and was released to the Russian market.

Nanimator

According to the company, the service optimizes the accuracy of the assessment due to the size of the data set and the principles of Big Data (big data). He is also able to highlight patterns that are invisible to a person, important for specific vacancies.

It is clear that as of January 2024, artificial intelligence is trying to introduce recruiting services into all services, but often this is not justified at all. For example, a large HR platform has created a recommendation system that learns from the opinion of a particular recruiter and immediately acquires him with biases. In other words, the original data of this system are based on the opinion of a person with certain preferences. He may not like young women who graduated from Moscow State University, or men with a short haircut. In our case, the real data of the company for which we train the model is scaled. For this reason, we do not use the cloud, only an on-premium solution - companies do not want (and should not) share their employees' data with the whole world. As a result, we get a set of data. It significantly optimizes the time for the selection of resumes, qualitatively improves it and optimizes the closure of positions by choosing the most suitable candidates. told Egor Muravyov, SRO of IT company Proscom

Nanimator

A nanimator is a service based on an NLP model and algorithmic machine learning:

• NLP model (Natural Language Processing). It is a machine learning technology that makes it possible to interpret and "understand" human language;
• algorithms look for patterns of successful and unsuccessful resumes. It is important to note that the marker of success and failure for training the model is set by the person who teaches it. So he teaches the machine to distinguish good from bad, showing examples of suitable and inappropriate resumes.

The service is able to determine: the presence or absence of key skills and experience, job specifications, the relevance of the experience of successful candidates. The model takes into account the requirements specified in the job advertisement and compares them with the information in the summary. This allows you to identify the correspondence between the expectations of the employer and the information provided by the candidate.

Nanimator

The nanimator is trained on two sets of data - on the resume of applicants who came up and did not come up. Algorithms look to the experience of successful candidates, match their characteristics and traits with new CVs to highlight the commonalities of eligible employees. After analyzing the resume, the service compiles candidate ratings for each vacancy. He also completely excludes the human factor in the processing of the resume - he is guided only by the skills and experience of the candidate.

Nanimator has already loaded several models trained in the most common IT vacancies in Russia: project manager, system analyst, product designer and others.

To be most effective, organizations are encouraged to create and train models based on their own data. At the same time, the Proscom team will help train new models: it will independently select resumes for training algorithms, taking into account the necessary key skills, and enrich them with real data from the organization. As a result, the customer company will receive a model trained on its own data, and not on the opinion of individual recruiters.

For one year, the company investigated the HR processes of one large bank (with 10,000 employees): selection, interviews, hiring, onboarding, training and direct work.

Nanimator

As a result, it was found that the organization employs 2,000 line employees a year and spends resources on their training and work. However, 30% of these workers quit in the first 3 months. So the bank loses up to 50,000,000 rubles a year. The animator will reduce losses on hiring and maintenance of personnel, since it will reduce the number of layoffs at the initiative of the employee due to more relevant and targeted selection. The service also optimizes the cost of hiring - HR specialists spend less time viewing resumes and quickly close vacancies.

In the near future, Proscom plans to launch a version for applicants - with the help of Nanimator, it will be possible to create an "ideal" resume. An array of accumulated impersonal data (models and user requests from the b2b section), market analytics and partner competencies will allow you to find out what skills, qualities and experience are needed to improve your resume and develop your career.

You can purchase the product under an on-premise license for installation on your own servers. With off-the-shelf models trained on the most common jobs, recruiters can test the service and understand how it works. But in order to get the maximum effect, recruiters need to create their own unique models based on company data.

The cost consists of the following items:

• installation;
• integration;
• license: 12 months or indefinite. In the case of a perpetual license, you must purchase a service support option in order to receive updates and third-line support for 12 months;
• assistance in training your own models;
• individual modifications in accordance with the client's requests.