AI Recruitment Assistant

Problem

A recurring hiring bottleneck is that resume emails arrive scattered, formats vary, job matching is manual, and candidate replies rarely get tracked. This skill turns IMAP email intake, resume parsing, job routing, AI scoring, question emails, and reply tracking into a runnable operational pipeline, focusing on the repetitive work between receiving a resume and completing an initial screening loop.

How It Works

The skill defaults to the WorkBuddy platform LLM, but can also use a custom external API such as DeepSeek, OpenAI, or Moonshot. The main steps are:

• Resume email intake: poll IMAP every 5 minutes and accept PDF/DOCX
• Resume parsing: extract name, contact details, and work history into plain text
• Job routing: match each candidate to the most suitable role
• AI scoring: score candidates across skills, experience, projects, education, and potential on a 100-point scale
• Question generation: produce 3-5 personalized questions that reference the resume
• Sending and tracking: send question emails via SMTP and check replies every 10 minutes
• Evaluation and timeout handling: cross-check replies, detect proxy answers, and resend after 3 days if no reply
• High-potential routing: send a recommendation email when the score is ≥80

Boundaries

This is best for hiring operations that already maintain a stable resume mailbox, role descriptions, and a recommendation inbox. It does not replace interview decisions, and parsed resume fields may still need human validation. External LLM availability, SMTP/IMAP credentials, and candidate inbox deliverability must be configured and tested in the admin portal.