Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Let's Get You Hired: A Job Seeker's Perspective on Multi-Agent Recruitment Systems for Explaining Hiring Decisions

Aditya Bhattacharya, Katrien Verbert

TL;DR

This paper tackles the opacity of hiring decisions by introducing a user-centered, multi-agent xCUI that uses LLMs to provide actionable, explainable feedback to job seekers. Through an iterative design process and qualitative evaluation with 20 participants, the system demonstrates higher perceived actionability, trust, and fairness compared with conventional recruitment methods. Contributions include a theoretical framework for iterative, user-centric MAS design in Explainable AI, a tangible open-source multi-agent recruitment chatbot, and empirical insights that guide design implications across domains. The findings suggest practical impact in improving candidate guidance, reducing ambiguity in hiring decisions, and informing broader adoption of user-aligned, multi-agent explainable AI systems.

Abstract

During job recruitment, traditional applicant selection methods often lack transparency. Candidates are rarely given sufficient justifications for recruiting decisions, whether they are made manually by human recruiters or through the use of black-box Applicant Tracking Systems (ATS). To address this problem, our work introduces a multi-agent AI system that uses Large Language Models (LLMs) to guide job seekers during the recruitment process. Using an iterative user-centric design approach, we first conducted a two-phased exploratory study with four active job seekers to inform the design and development of the system. Subsequently, we conducted an in-depth, qualitative user study with 20 active job seekers through individual one-to-one interviews to evaluate the developed prototype. The results of our evaluation demonstrate that participants perceived our multi-agent recruitment system as significantly more actionable, trustworthy, and fair compared to traditional methods. Our study further helped us uncover in-depth insights into factors contributing to these perceived user experiences. Drawing from these insights, we offer broader design implications for building user-aligned, multi-agent explainable AI systems across diverse domains.

Let's Get You Hired: A Job Seeker's Perspective on Multi-Agent Recruitment Systems for Explaining Hiring Decisions

TL;DR

This paper tackles the opacity of hiring decisions by introducing a user-centered, multi-agent xCUI that uses LLMs to provide actionable, explainable feedback to job seekers. Through an iterative design process and qualitative evaluation with 20 participants, the system demonstrates higher perceived actionability, trust, and fairness compared with conventional recruitment methods. Contributions include a theoretical framework for iterative, user-centric MAS design in Explainable AI, a tangible open-source multi-agent recruitment chatbot, and empirical insights that guide design implications across domains. The findings suggest practical impact in improving candidate guidance, reducing ambiguity in hiring decisions, and informing broader adoption of user-aligned, multi-agent explainable AI systems.

Abstract

During job recruitment, traditional applicant selection methods often lack transparency. Candidates are rarely given sufficient justifications for recruiting decisions, whether they are made manually by human recruiters or through the use of black-box Applicant Tracking Systems (ATS). To address this problem, our work introduces a multi-agent AI system that uses Large Language Models (LLMs) to guide job seekers during the recruitment process. Using an iterative user-centric design approach, we first conducted a two-phased exploratory study with four active job seekers to inform the design and development of the system. Subsequently, we conducted an in-depth, qualitative user study with 20 active job seekers through individual one-to-one interviews to evaluate the developed prototype. The results of our evaluation demonstrate that participants perceived our multi-agent recruitment system as significantly more actionable, trustworthy, and fair compared to traditional methods. Our study further helped us uncover in-depth insights into factors contributing to these perceived user experiences. Drawing from these insights, we offer broader design implications for building user-aligned, multi-agent explainable AI systems across diverse domains.

Paper Structure

This paper contains 43 sections, 6 figures, 1 table.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. AI Agents
  4. User Interfaces for Conversational Explanation
  5. Application Implementation: Multi Agentic Recruitment System
  6. Usage Scenario
  7. Exploratory User Study
  8. User Requirements
  9. Multi-Agentic System Design
  10. Agent Roles and Responsibilities
  11. Architectural Design Pattern
  12. Agent Tools
  13. Agent Memory Management
  14. High-Fidelity Prototype
  15. Qualitative User Study
  16. ...and 28 more sections

Figures (6)

  • Figure 1: Iterative user-centric design and development approach for our multi-agentic recruitment system.
  • Figure 2: Screenshot of our multi-agent xCUI application with the different UI components marked: (a) UI control to upload the resume and job requirement documents, (b) this component illustrates the matching score with summarised explanations, (c) configuration controls to alter the display of certain UI components, (d) personalised agent introduction, (e) quick questions to help users start the chatbot conversation and (f) agentic AI responses presented in a simplified format with necessary visual aids along with input text box for user queries.
  • Figure 3: Study flow of our qualitative user study with 20 active job seekers.
  • Figure 4: Box plots showing an increase in the perceived actionability of our multi-agentic AI feedback compared to conventional methods
  • Figure 5: Box plots showing an increase in the perceived trust of our multi-agentic AI feedback compared to conventional methods
  • ...and 1 more figures