Recruitment Assistant V1 Agent

Agent Description:

The Recruitment Assistant V1 Agent is designed to automate the initial stages of candidate evaluation by analyzing professional CVs and matching them against specific job requirements. It enables HR teams to efficiently process large volumes of applications by extracting key qualifications and quantifying how well a candidate fits a designated role.

Purpose and Components
  • Purpose: This template streamlines the screening workflow by converting unstructured documents (PDF, DOCX, Text) into structured intelligence. It identifies missing or strong skills, validates data consistency, and uses a weighted scoring mechanism to categorize applicants. A unique feature of this agent is its ability to autonomously enrich the evaluation context by retrieving missing job description details via RAG (Retrieval-Augmented Generation) if the initial input is incomplete.

    The agent improves recruitment accuracy and speed by:

    • Extracting and normalizing core professional fields from unstructured CVs.

    • Assigning confidence scores to extracted data to ensure high-quality analysis.

    • Performing semantic skill matching using embeddings and keyword similarity.

    • Automatically retrieving missing job description requirements using the "job_description_rag" tool.

    • Generating holistic, weighted scores based on skills, experience, and profile consistency.

  • Components:

    • CV Extractor: Parses unstructured resume files into structured JSON and validates data completeness.

    • Skill Evaluator: Compares applicant qualifications against job requirements and identifies specific skill gaps.

    • Final Scorer: Aggregates all evaluative data to produce a weighted fit score and categorical label (Excellent, Moderate, or Low Fit).

Supported Capabilities

  • Multi-format CV parsing (PDF/DOCX/Text)

  • Data normalization for Name, Skills, Experience, and Education

  • Extraction confidence scoring (0.0–1.0)

  • Semantic skill match analysis via embeddings

  • Missing skill detection and reporting

  • Automated RAG-based job description retrieval for incomplete role data

  • Weighted holistic scoring: 40% Skills, 30% Experience, 20% Consistency, 10% Engagement

  • Final fit justification and executive summary generation

LLM Used

  • OPENAI GPT_4O_MINI

    Note: To learn more about the LLM and to modify its behavior, refer to the Configuring LLM settings section.

Sub-Agents

1. CV Extractor

  • Role:CV Extractor

  • Scope:Convert unstructured CVs into structured JSON with confidence scoring.

  • Description:Extracts professional fields and validates them for consistency. Assigns a confidence score where data clarity is high (>0.7) or flags records as low-confidence if key sections like experience or education are vague.

  • LLM Used: Default (Inherits from parent agent).

2. Skill Evaluator

  • Role:Skill Evaluator

  • Scope:Evaluate candidate skills and experience against job requirements.

  • Description:Compares CV qualifications to job descriptions using keyword similarity. If the JD is incomplete, it invokes the job_description_rag tool to fetch requirements based on the role title before outputting a match report.

  • LLM Used: Default (Inherits from parent agent).

3. Final Scorer

  • Role:Final Scorer

  • Scope:Generate a holistic, weighted evaluation of the candidate.

  • Description:Aggregates all previous data points to compute a final score. Labels candidates as Excellent, Moderate, or Low Fit and provides a summary with a detailed logic-based justification.

  • LLM Used: Default (Inherits from parent agent).

Tools Used:

  • MCP Client Tool- Retrieves role-specific requirements and responsibilities when job descriptions are missing or incomplete.

Note: For details on modifying the Tools, refer Tools Library section.
Agent Workflow Behavior Summary

  1. Input: The user provides a candidate's CV and a target job title or description.

  2. Extraction: The CV Extractor normalizes the resume data; if the confidence is below 0.7, the record is flagged for potential manual review.

  3. Context Enrichment: If the job requirements provided are vague, the Skill Evaluator calls the RAG tool to pull industry-standard requirements for that specific role title.

  4. Skill Assessment: The Evaluator computes a score (0–1); scores below 0.75 are flagged for significant skill gaps.

  5. Weighted Scoring: The Final Scorer applies the percentage weights (for example, 40% for skills) to generate the final fit percentage.

  6. Final Output: A candidate evaluation report with a fit label and reasoning.

Sample Questions:

  • Analyze the attached CV for the Lead Data Scientist position.

  • Compare John Doe's resume against our current requirements for a Java Developer.