Where recruitment time disappears
A typical Australian recruitment consultant spends 60–70% of their week on three things: screening unsuitable candidates out of the funnel, writing job descriptions that sound generic, and answering the same 12 candidate questions over and over. None of these are billable activities. All of them are perfect for a custom GPT.
We build AI assistants that handle the screening, drafting, and FAQ workload — so consultants can spend their time on client relationships, candidate placement, and the complex matching judgement calls that AI can't do.
What we build for recruitment agencies
Candidate Screening Assistant
CV uploaded, candidate gets 6 role-specific questions via SMS or email. Answers analysed against the role brief. Strong fits go to the consultant; weak fits get a polite, branded rejection. Consultant only sees the top 30%.
JD Drafting Bot
Consultant inputs the client name, the role, and 4 bullet points from the briefing call. Output: a 380-word JD in the firm's house style with the right tone for the seniority level.
Candidate Enquiry Bot
Sits on your job ads. Answers 'is this hybrid or fully remote?', 'what's the salary range?', 'when do interviews start?' — using your live data from JobAdder, Bullhorn, or FastTrack.
Reference Checking Helper
Drafts reference-check question sets specific to the role and seniority. Captures responses via a structured form. Generates the consolidated reference report for the client.
ATS and CRM integrations
- JobAdder — full REST API support, candidate sync, vacancy data
- Bullhorn — SOAP and REST API integration
- FastTrack — read-only ATS integration
- Vincere — REST API support
- SEEK Talent Search — search and shortlist integration
What recruitment firms tell us
A 22-consultant agency in Sydney's tech recruitment space reduced average time-from-application-to-shortlist from 6 days to 1.5 days using an automated screening GPT. Conversion from application to consultant call rose from 8% (manual screening, slow) to 14% (instant screening). A blue-collar staffing firm in Western Australia using the JD drafting bot saved an estimated 280 consultant hours per quarter that previously went to writing role descriptions.
Bias-aware by design. Screening prompts are reviewed by an HR specialist before deployment to remove indirectly discriminatory criteria (e.g., age proxies via grad year, gender-coded language). We log every screening decision so you can audit fairness if a candidate ever raises a concern.
Frequently asked questions
Won't candidates feel insulted being screened by AI?
Not when it's done right. The GPT is transparent about being an AI assistant, the questions are role-relevant (not pop psychology), and the response time is faster than a human would manage. Candidate satisfaction surveys we've run show higher scores for AI-assisted screening than for the slow human equivalent — speed wins.
How does it integrate with JobAdder?
JobAdder offers a comprehensive REST API. We connect via OAuth, pull active vacancies and candidate records, and write screening results back as candidate notes or custom field updates. Setup is typically 3–4 hours of integration work.
What about the privacy of candidate CV data?
Australian Privacy Act 1988 applies. We deploy in Australian-region cloud infrastructure, sign DPAs that exclude training on candidate data, and provide a candidate-facing privacy disclosure that the AI screening is happening. We've successfully deployed under standard recruitment industry privacy frameworks for 8+ Australian agencies.
Can it handle volume — 500+ applications per role?
Yes. The screening throughput is essentially unlimited (the constraint is your candidate-response rate, not the GPT's processing capacity). We've handled blue-collar campaigns with 1,200 applicants in 48 hours, fully screened and ranked, with the consultant seeing only the top 90 fits.
Ready to build your custom GPT?
Get a free 30-minute scoping call. We'll map your use case, data sources, and ROI before you commit.
Start the Conversation