TRACKED CALLS / 24H1,284+12.4%SMS RECOVERIES / WK287+8.1%DEMAND REACTIVATION RESPONSE RATE41.0%+3.2ppAU + NZ CLIENTS LIVE512+11AVG QUOTE RESPONSE47 MIN-9 minISO 27001CERTIFIEDTRACKED CALLS / 24H1,284+12.4%SMS RECOVERIES / WK287+8.1%DEMAND REACTIVATION RESPONSE RATE41.0%+3.2ppAU + NZ CLIENTS LIVE512+11AVG QUOTE RESPONSE47 MIN-9 minISO 27001CERTIFIED
LIVESANDBOXAI BUILDSCASE STUDYSYDNEY · MELBOURNE · BRISBANE · PERTH · AUCKLANDMon to Fri · 9am to 5pm AEST

Sandbox · What we deliver for operators who need a custom AI build

// Interviewer Portal · client build

Mr Jobs Recruitment Engine

Interviewer Portal · client build

Mr Jobs Recruitment Engine

problem

The problem. A recruitment operator has 1,000 candidates to review. They need to shortlist, score, and send interview invites at scale - without reading every CV manually, without a team of reviewers, and without dropping confidence in who gets through. The old approach is a spreadsheet, a hiring manager's gut, and a weekend of email. It does not scale. The real problem is not the volume. It is doing bulk candidate review and automated invite dispatch while keeping the confidence level high enough that the shortlist is actually worth interviewing.

What we built. A recruitment platform that processes candidates in bulk at whatever volume the operator needs. Document AI parses every resume into structured data the moment it arrives. Each candidate is scored across six weighted dimensions - skills match, experience relevance, tenure stability, recency, qualifications, overall strength - with explicit AI rationale and risk flags per decision, not a black-box score. High-confidence candidates are automatically sent an AI interview link. A Vapi voice agent, voiced by ElevenLabs, conducts the screening interview autonomously on the candidate's schedule. When the interview completes, the transcript and responses come back into the platform, the candidate is re-scored against the role, and the updated shortlist surfaces the top candidates to the operator. Low-confidence candidates are routed for human review. No operator time spent until the ranked shortlist lands in their queue. Every decision is auditable. Built on Vertex AI with model selection by task complexity.

What it produces. 1,000 candidates reviewed, AI-interviewed, re-scored, and ranked without a single manual triage step. High-confidence candidates complete a full voice screening interview before a human sees them. Hiring managers receive a ranked shortlist of interviewed candidates with scoring rationale per dimension, not a stack of PDFs. Every interview recorded, every score auditable, every hire defensible.

Where the value flows.

  • +Scale: 1,000+ candidates processed in bulk. No manual inbox triage.
  • +Scoring: six-dimension AI scoring with explicit rationale and risk flags on every resume.
  • +Interview dispatch: high-confidence candidates receive an automated AI interview link. No operator action required.
  • +AI interview: Vapi voice agent with ElevenLabs voice conducts the screening interview autonomously on the candidate's schedule.
  • +Re-scoring: interview responses feed back into the platform. Candidate score updated. Top candidates surface to the operator.
  • +Control: low-confidence candidates routed for human review. Nothing slips through unscored.
  • +Compliance: every interview recorded, every decision audit-trailed.
Mr Jobs AI candidate scoring panel showing a 72 out of 100 score with 90 percent AI confidence, broken down across six weighted dimensions: skills match, experience relevance, tenure stability, recency, qualifications, and overall strength, with full AI rationale and risk flags listed.
AI candidate scoring. Every applicant scored across six weighted dimensions with explicit AI rationale and auditable risk flags (missing certs, tenure gaps). No black-box decisions. All data shown is dummy.

BETA. Working prototype built in 1 week. Interviewer portal in active development.

‹ Back to all Sandbox builds

Want one like this in your business?

Every Sandbox build started as one operator's bottleneck. Tell us yours.

A 30-minute diagnose call, no charge. We tell you whether it is a 48-hour build or something bigger, before you commit a dollar.