Screen Inbound Resumes and Shortlist into an Airtable Tracker

This workflow turns a noisy hiring inbox into a clean, ranked shortlist with zero manual triage. It watches a Gmail label (e.g. `careers/inbound`) for new applications, extracts the attached resume and cover note, and asks OpenAI to score the candidate against a structured rubric you define for the role — years of relevant experience, must-have skills, location/work authorization fit, and a short justification. Every candidate is written to an Airtable base as a new row with the AI score, tier (Strong / Maybe / Pass), parsed contact details, and a one-paragraph summary, so recruiters open Airtable already looking at a sorted shortlist instead of an inbox.

Use it when a single role or careers alias is pulling in more resumes than your team can read fairly and quickly, and you want a consistent, auditable first-pass screen instead of gut-feel inbox skimming. It is ideal for lean teams without an expensive ATS, for high-volume reqs (support, sales, ops) where the top of funnel is the bottleneck, and any time you need every applicant logged in one place with a defensible score and reason attached. Pair it with a recruiter review step in Airtable — the AI ranks, humans decide.

The trigger fires when a new email lands under the configured Gmail label. The workflow downloads the resume attachment (PDF or DOCX) and pulls the email body, then sends both to OpenAI with a role-specific scoring prompt that returns structured JSON — name, email, score 0-100, tier, matched skills, and a summary. A logic step normalizes the parsed fields and maps the score to a tier band. Finally the candidate record is upserted into the Airtable applicant tracker (deduped on email), tagged with the role and a timestamp, ready for recruiter review. Because this runs on Agent Hive's Sim engine, the Gmail watch, AI scoring, and Airtable write all execute live in your colony — not as a mockup.