Approach
N-square AI built a Custom Recruiting AI Agent System with four specialized agents (Job Posting, Candidate Sourcing & Scoring, Outreach & Qualification, Interview Coordination) orchestrated through a Human-in-the-Loop dashboard, automating high-volume tasks while ensuring recruiter control at every critical decision point.
100% control through a Human-in-the-Loop interface.
Reliability at scale across high-volume tasks.
Direct ROI by removing operational drag and increasing candidate throughput.
Overview
DKF Recruitment is an award-winning, pioneering recruitment company specializing in The Built Environment. Founded in 2014 by owner and managing director Dominic Fagan, DKF has established itself as a market leader in recruitment and labour solutions for asbestos removal, asbestos management and treatment, demolition, soft-strip, and land remediation. Trading nationally from their head office in Altrincham, South Manchester, DKF provides services supported by many years of experience, proven successes, and exemplary industry compliance.
DKF partnered with TrueHorizon AI to embed an advanced, custom AI Agent into its talent acquisition workflows. The goal was to transform recruiting from a series of fragmented, manual tasks into an automated, high-velocity pipeline while ensuring human oversight at every critical decision point.
The solution is a Custom Recruiting AI Agent System that automates the full lifecycle—from job creation and candidate sourcing to qualification and interview coordination.
The Problem: Recruiting Bottlenecks and Manual Drag
Recruiting teams often struggle with bottlenecks caused by repetitive, low-value administrative work, leading to missed opportunities and a slower time-to-hire.
Key challenges for DKF included:
Manual job posting across multiple boards (LinkedIn, Indeed, etc.).
Time-consuming candidate sourcing and resume review.
Inconsistent or delayed candidate outreach and qualification.
Interview scheduling requiring continuous back-and-forth communication.
A lack of high-level performance visibility for senior leadership.
Recruiters were spending valuable time on tasks that required machine speed, not human judgment, throttling the overall pipeline.
The Ask: Agentic Recruiting with Human Control
DKF's vision was for an AI agent that could handle the high-volume, repetitive work, acting as an always-on assistant, but with a safety-first architecture.
The platform needed an agent that:
Works with human recruiters, not in place of them (HITL model).
Provides measurable KPIs for leadership (e.g., time-to-hire, cost-per-hire savings).
Is flexible, integrating seamlessly with their existing Applicant Tracking System (ATS).
Could perform complex, multi-step actions (e.g., personalized outreach, conversational qualification).
The Solution: A Modular, Four-Agent Architecture
TrueHorizon AI designed a modular, scalable AI Agentic System, built with preferred tools like n8n for workflow orchestration and an LLM API (e.g., Grok) for natural language tasks.
The system is composed of four specialized agents, all connected through a central Human-in-the-Loop (HITL) Dashboard:
1. Job Posting Agent
Action: Triggers on new job req in the ATS. LLM generates a job post draft based on the description and company templates.
HITL: Recruiter reviews and approves or edits the draft in the UI.
Output: Automated posting to specified job boards (LinkedIn, OCC, Empresa MX, Indeed).
2. Candidate Sourcing & Scoring Agent
Action: Scrapes and sources candidate profiles from external sources (e.g., LinkedIn API). LLM scores each profile (e.g., 0-100 fit score) against job criteria.
HITL: Recruiter reviews the queue and performs a bulk approve or reject action.
Output: Imports qualified candidates and their scores into the ATS.
3. Outreach & Qualification Agent
Action: Initiates personalized outreach sequences (email/SMS) and engages in qualification dialogues (asking/answering questions via LLM).
HITL: Recruiter can monitor conversations in real-time and intervene or override responses if necessary.
Output: Hands off a warm, qualified lead with a conversation summary.
4. Interview Coordination Agent
Action: Integrates with calendar APIs (e.g., Google/Outlook) to automatically coordinate and schedule interviews. After the interview, it summarizes feedback from recruiter notes.
HITL: Recruiters can make manual adjustments to schedules or feedback.
Output: Updated ATS records reflecting interview status and feedback.
The system is designed to:
Automate repetitive, high-volume tasks (e.g., job posting, screening, scheduling).
Provide LLM-based candidate scoring for rapid shortlisting.
Enable Human-in-the-Loop (HITL) controls for all major decisions (e.g., approving a job post, rejecting a candidate).
Reduce time-to-hire and free up recruiter hours for high-value engagement and relationship-building.
By deploying this agentic system, DKF anticipates achieving the following key outcomes:
Job Posting Agent – Automated job posting to multiple boards with LLM-generated drafts and recruiter approval
Candidate Sourcing & Scoring Agent – Intelligent candidate sourcing with AI-powered fit scoring
Outreach & Qualification Agent – Personalized candidate engagement with conversational qualification
Interview Coordination Agent – Automated interview scheduling and feedback summarization
Human-in-the-Loop Dashboard – Central control interface for recruiter oversight and intervention
This deployment shifts the recruiter's role from administrative processing to strategic talent engagement:
Time Savings: Converting highly manual, repetitive work into automated processes.
Increased Pipeline - Boosting the number of qualified candidates sourced per role.
Efficiency Gains - Reducing time-to-screen (resume review to shortlist) and overall time-to-hire.
Cost Reduction - Lowering the cost-per-hire by cutting manual hours and reducing reliance on external resources.
Recruiter Focus - Freeing up recruiter hours for high-value activities like relationship-building and core interviewing.
Visibility - Providing senior leaders with a real-time KPI dashboard of the entire pipeline and agent efficiency.
This project is not about simple automation or a standalone chatbot.
It is a production-grade AI workforce built specifically for the recruiting lifecycle, designed to deliver:
100% control through a Human-in-the-Loop interface.
Reliability at scale across high-volume tasks.
Direct ROI by removing operational drag and increasing candidate throughput.
DKF is building a future where every recruiter can act as a high-powered, high-velocity talent partner.