NLP BASED AUTOMATED INTERVIEW AND CANDIDATE ASSESSMENT PLATFORM FOR HR MANAGEMENT

Abstract

The present invention discloses an applicant tracking system (ATS)integrated with a large language model (LLM) to enhance recruitment processes by automating key tasks and improving accuracy. Traditional ATS platforms rely on keyword matching and manual workflows, often leading to inefficiencies and bias. This system leverages advanced natural language processing to screen resumes contextually, ensuring a more accurate match between candidates and job requirements. The ATS also automates interview scheduling, job distribution, and candidate communication, significantly reducing administrative workload. The system integrates customisable workflows, allowing recruiters to tailor the hiring process to specific job roles. Collaboration tools,along with real- time reporting and analytics, enhance team coordination and enable data-driven decision-making. Additionally, AI-driven processes help reduce unconscious hiring bias, promoting more equitable recruitment. Overall, the inventionprovides a more efficient, accurate, and user-friendly solution to modern recruitment challenges, significantly advancing traditionalATS capabilities.

Specification

Description:NLP BASED AUTOMATED INTERVIEW AND CANDIDATE ASSESSMENT PLATFORM FOR HR MANAGEMENT

Technical Field of the Invention:

[0001] The present invention relates to the field of recruitment technology, specifically applicant tracking systems (ATS)
5 enhanced with large language models (LLM) for automating and streamlining the hiring process. It is particularly focused on improving candidate sourcing, reducing time to hire, and minimising hiring bias.

10 Background of the Invention:

[0002] The recruitment and hiring landscape have undergone significant transformation over the past few decades, driven by advancements in technology. Traditional hiring processes were
15 once dominated by manual procedures, where job openings were posted in newspapers and employers received hard-copy applications. This method was time-consuming, costly, and prone to human error. As businesses scaled, so did the need for more efficient ways to manage the influx of applicants, leading to the
20 development of applicant tracking systems (ATS). These early systems, designed to automate aspects of recruitment, allowed organisations to manage job postings and store candidate data such as resumes and application status. While a step forward, these ATS solutions primarily focused on administrative functions
25 rather than improving the candidate evaluation process.

[0003] Presently, many ATS platforms incorporate more advanced features like resume parsing, keyword searches, and interview scheduling. However, they still suffer from critical limitations.



Existing ATS technology often relies on basic algorithms that may filter out qualified candidates due to the use of rigid keyword matching. As a result, applicants who do not perfectly align with predefined criteria may be unfairly excluded, despite having the
5 relevant skills and experience. Moreover, many ATS systems are cumbersome to use, offering limited customisation and collaboration tools for recruiters. Communication with candidates often remains slow and inefficient, leading to poor candidate experience and missed opportunities for businesses to secure top
10 talent. Another major drawback of current systems is their inability to effectively reduce hiring bias, as they lack sophisticated AI tools to identify and address unconscious bias during the screening process.

15 The Patent titled "Methods and systems of employment candidate data management" discloses A method and system for managing employment candidate data utilizes a pre-screen application to enable pre- screening of employment candidates who log on to a website and
20 are prompted to enter additional information and schedule pre- employment tests via the website. The method and system for managing employment candidate data also uses a tracking and testing application to handle screening of candidates' paper-based applications, administration of interviews, hiring pool
25 management, job offer support, and scheduling of training, and is integrated with an applicant tracking system. A resume tracking application of the method and system for managing employment candidate data automatically identifies and merges duplicate resume information to create a single unique resume in the



tracking system. However, the present invention introduces AI- driven automation through large language models, addressing modern recruitment challenges more comprehensively than the earlier invention.
5
[0005] To address these limitations, the field of recruitment technology needs a more intelligent and flexible solution that can better match candidates to roles based on comprehensive analysis rather than rigid keyword searches. Moreover, systems should
10 offer enhanced collaboration capabilities for recruitment teams, customisable workflows, and real-time messaging to improve communication with candidates. Reducing time-to-hire while enhancing candidate quality is essential, as organisations seek to streamline their recruitment strategies to remain competitive.
15 There is also a growing need for AI-driven analytics and reporting tools that can help HR teams measure the effectiveness of their hiring efforts and make data-driven decisions. Finally, improving the candidate experience through automated, timely communication and reducing hiring bias are critical elements that
20 current ATS platforms must address to evolve with the demands of modern recruitment.

Summary of the invention:

25 [0006] The present invention discloses an applicant tracking system (ATS) with large language model (LLM) integration, addressing the significant limitations of conventional recruitment technologies. Traditional ATS platforms are often burdened by inefficiencies such as reliance on basic keyword matching, which can lead to



inaccurate candidate filtering. These systems typically struggle with manual processes for resume screening, interview scheduling, and candidate communication, resulting in prolonged time-to-hire, increased costs, and a less satisfactory candidate
5 experience. The rigidity and lack of advanced analytical capabilities in existing ATS solutions hinder their effectiveness in adapting to the dynamic needs of modern recruitment.
[0007] The present invention utilises cutting-edge LLM technology to transform the resume screening process. By employing
10 advanced natural language processing capabilities, the system goes beyond simple keyword matching to understand the context and relevance of candidate qualifications. This comprehensive analysis improves the accuracy of candidate matching, ensuring that highly qualified applicants are not overlooked due to
15 limitations in keyword-based filtering. This innovation addresses one of the major drawbacks of traditional ATS platforms, which often miss out on candidates who may have the right skills but do not use the exact keywords specified in job descriptions.

20 [0008] Additionally, the invention streamlines interview scheduling through automation, reducing the time and effort required to coordinate between candidates and recruiters. The system manages scheduling conflicts, sends reminders, and handles rescheduling requests with minimal manual intervention. This
25 automation not only enhances efficiency but also minimises the potential for scheduling errors, which can be a common issue in manual processes.



[0009] The present invention further introduces advanced collaboration tools that foster better teamwork among HR professionals. Features such as shared dashboards, collaborative notes, and approval workflows enable recruitment teams to work
5 together more effectively, ensuring a more organised and cohesive approach to hiring. This collaborative capability addresses the limitations of traditional systems, which often lack integrated tools for team coordination and communication. Furthermore, customisable workflows and communication methods enhance the
10 system's adaptability to various recruitment needs. Recruiters can tailor the ATS to fit specific job roles and organisational requirements, allowing for greater flexibility and precision in the recruitment process. Real-time messaging capabilities ensure that candidates receive timely updates and communication, improving
15 their overall experience and engagement.

[0010] The invention also addresses the challenge of job distribution and candidate aggregation by automating these processes. The ATS posts job openings across multiple platforms
20 and consolidates applications from various sources, thereby broadening the talent pool and increasing the reach of job postings. This feature overcomes the limitations of manual job distribution and aggregation, which can be time-consuming and less effective.
25
[0011] Furthermore, the system's use of AI-driven processes helps to reduce unconscious hiring bias. By implementing advanced algorithms for candidate evaluation, the ATS promotes a more equitable and fair recruitment process, countering one of the



major concerns in traditional hiring methods. The comprehensive reporting and analytics provided by the system offer valuable insights into recruitment strategies, enabling HR teams to make informed, data-driven decisions.
5
Brief description of the drawings:


[0012] The foregoing and other features of embodiments will become more apparent from the following detailed description of
10 embodiments when read in conjunction with the accompanying drawings.

[0013] FIG 1 illustrates a system architecture for an applicant tracking system (ATS) integrated with a large language model.
15

Detailed description of the invention:

[0014] The present invention pertains to an applicant tracking
20 system (ATS) integrated with a large language model (LLM), designed to address the limitations of conventional recruitment technologies. This advanced ATS incorporates a range of novel features and components that significantly enhance the recruitment process, making it more efficient, accurate, and user-
25 friendly. Below is a detailed description of the invention and its components, structured according to different embodiments.

[0015] Figure 1 illustrates a system architecture for an applicant tracking system (ATS) integrated with a large language model
30 (LLM), designed to streamline and optimise the recruitment



process. The system's core component is the processing unit (102), which coordinates all critical functions, including data input, processing, and communication across different modules. This unit plays a central role by analysing recruitment data, facilitating
5 decision-making, and driving automation in tasks such as resume screening, interview scheduling, and candidate communication.

[0016] Data input is managed through the input unit (101), which collects information from various sources, including candidate
10 resumes, job descriptions, and external recruitment platforms. The input unit feeds this data into the system, ensuring that the ATS can handle multiple types of input and formats, enhancing the system's flexibility and scalability. Paired with the user interface module (103), recruiters can seamlessly interact with the system,
15 configure workflows, and access candidate information. This interface is user-friendly, ensuring that HR professionals can manage the recruitment process efficiently without requiring technical expertise.

20 [0017] The database module (104) is responsible for securely storing all recruitment-related data, such as candidate profiles, resumes, and communication logs. It ensures quick retrieval of data and maintains the integrity of stored information, allowing the processing unit to access and analyse relevant data effectively.
25 The network interface module (105) automates job distribution and candidate aggregation by linking the ATS to external platforms like job boards and social media, ensuring a wider reach for job postings and streamlining the application collection process.



[0018] Finally, the output unit (106) delivers critical outputs such as ranked candidate lists, interview schedules, and comprehensive recruitment reports. By providing organised and relevant insights, this module helps HR teams make data-driven decisions that
5 optimise the recruitment process. Overall, this invention integrates automation, advanced language processing, and comprehensive data management to offer a powerful, user- friendly, and efficient ATS solution that addresses the limitations of traditional recruitment technologies.
10
[0019] In one embodiment of the invention, the ATS utilises advanced large language model (LLM) technology to revolutionise resume screening. Traditional ATS platforms often rely on keyword matching, which can overlook highly qualified candidates who may
15 not use the exact keywords specified in job descriptions. Incontrast, the LLM integration enables the system to understand the context and relevance of candidate qualifications beyond simple keyword matching. This embodiment involves an AI model that processes resumes through natural language processing
20 (NLP) techniques, analysing the content for contextual accuracy and relevance. As a result, the system can identify and rank candidates more effectively, improving the accuracy of candidate matching and ensuring that no suitable candidate is overlooked due to limitations in keyword-based filtering.
25
[0020] In another embodiment of the invention, the ATS streamlines the interview scheduling process through automation. Manual scheduling of interviews can be time-consuming and prone to errors, leading to inefficiencies and delays in the hiring process.



This embodiment of the invention introduces an automated scheduling module that manages interview coordination between candidates and recruiters. The system handles scheduling conflicts, sends out interview reminders, and processes
5 rescheduling requests with minimal manual intervention. This automation not only reduces the administrative burden on HR personnel but also minimises the potential for scheduling errors, enhancing overall efficiency in the interview process.

10 [0021] Yet in another embodiment of the invention, advanced collaboration tools are integrated to improve teamwork among HR professionals. Traditional ATS systems often lack robust tools for team coordination and communication, leading to fragmented and inefficient recruitment processes. This embodiment addresses this
15 issue by incorporating features such as shared dashboards, collaborative notes, and approval workflows. These tools allow HR teams to work together more effectively by providing a centralised platform for managing recruitment activities, sharing insights, and making collective decisions. This collaborative capability ensures
20 a more organised and cohesive approach to hiring, facilitating better communication and coordination within the recruitment team.

[0022] In another embodiment of the invention, the ATS includes 25 customisable workflows and communication methods to enhance adaptability to various recruitment needs. The system allows recruiters to tailor workflows to fit specific job roles and organisational requirements, offering greater flexibility and precision in the recruitment process. Customisable features



include the ability to adjust the stages of the hiring process, modify communication templates, and set specific criteria for candidate evaluation. This embodiment ensures that the ATS can be adapted to meet the unique needs of different organisations
5 and job roles, allowing for a more tailored and effective recruitment process.

[0023] In a further embodiment of the invention, job distribution and candidate aggregation are automated to broaden the talent
10 pool and increase the reach of job postings. Traditional methods of job distribution often involve manually posting job openings across various platforms and managing applications from multiple sources. This embodiment introduces an automated job distribution module that posts job openings across multiple job
15 boards and recruitment platforms simultaneously. The system also consolidates applications from various sources into a centralised platform, streamlining the application management process. This feature helps to overcome the limitations of manual job distribution and aggregation, expanding the reach of job postings
20 and improving the efficiency of candidate sourcing.


[0024] In yet another embodiment of the invention, the ATS incorporates AI-driven processes to reduce unconscious hiring
25 bias. Traditional recruitment methods can be influenced by biases that affect candidate evaluation and selection. This embodiment utilises advanced algorithms and AI-driven evaluation techniques to promote a more equitable and fair recruitment process. The system's AI model assesses candidates based on objective criteria
30 and relevant qualifications, reducing the impact of unconscious



bias on hiring decisions. Additionally, the system provides comprehensive reporting and analytics on recruitment practices, offering insights into potential biases and helping HR teams implement more inclusive hiring strategies.
5
[0025] In another embodiment of the invention, the ATS offers comprehensive data-led recruitment reporting and analytics. Traditional ATS platforms often lack detailed reporting capabilities, making it challenging for HR teams to evaluate the effectiveness
10 of their recruitment strategies. This embodiment addresses this limitation by providing advanced analytics and reporting features that offer valuable insights into recruitment performance. The system generates detailed reports on metrics such as time to hire, cost per hire, and candidate quality, enabling HR teams to make
15 informed, data-driven decisions. This capability enhances the ability to evaluate and optimise recruitment strategies, leading to more effective and efficient hiring processes.

[0026] In summary, the present invention integrates large language
20 model technology with an applicant tracking system to address the shortcomings of traditional recruitment platforms. By incorporating advanced features such as automated resume screening, interview scheduling, collaboration tools, customisable workflows, job distribution, and AI-driven bias reduction, the
25 invention offers a comprehensive solution that enhances the efficiency, accuracy, and user experience of the recruitment process. The present invention significantly advances the capabilities of traditional ATS platforms by integrating LLM technology and automation features. It effectively addresses



existing drawbacks, resulting in a more efficient, accurate, and user-friendly recruitment solution that enhances the hiring process and improves the overall candidate experience.
, Claims:WE CLAIM:

1. A system for an Applicant Tracking System (ATS) with Large Language Model (LLM) Integration. The system (100)
5 comprising:
an input unit (101) configured to receive and process user inputs, including candidate data and job postings;
a processing unit (102) integrated with the system to execute large language model (LLM) algorithms for
10 candidate screening and other data processing tasks;
a user interface module (103) for handling interactions between the user and the system, including data entry and display of system outputs;
a database module (104) coupled with the processing unit,
15 for storing and managing candidate information, resumes, cover letters, and job postings;
a network interface module (105) for enabling communication between the system and external networks, such as job boards and company websites; and
20 an output unit (106) configured to display information and facilitate user interaction with the system, including screens and input devices.

2. The system (100) as claimed in claim 1, wherein the input unit
25 (101) includes data entry components such as keyboards and touchscreens for user interaction.



3. The system (100) as claimed in claim 1, wherein the processing unit (102) includes a dedicated processor for executing large language model (LLM) algorithms.

5 4. The system (100) as claimed in claim 1, wherein the database module (103) is a relational database management system (RDBMS) designed for structured data storage and retrieval.

5. The system (100) as claimed in claim 1, wherein the user
10 interface module (104) includes software components for managing user interactions and displaying system information.

6. The system (100) as claimed in claim 1, wherein the network interface module (105) supports both wired and wireless
15 communication protocols for connectivity with external networks.

7. The system (100) as claimed in claim 1, wherein the user interface device/output (106) includes high-resolution display
20 screens and input devices such as a mouse or stylus for user interaction.

Documents