Full applicant tracking: AI résumé parsing, candidate scoring, a careers page, pipeline, feedback and analytics — running inside your own Google account, with nothing to pay and nothing to lock you in.
Companies pay thousands a year for this. You won’t. · Open source · MIT
The complete loop — from requisition to offer debrief — running on Google Sheets, Drive, Gmail and Calendar you already have.
Titles, levels, salary bands, openings, priorities — with hiring manager and recruiter ownership.
A hosted apply page out of the box, plus CV-by-email: forward to careers@ and it's parsed automatically.
Gemini reads every CV into structured fields and scores it on a transparent rubric you can tune in plain English.
Kanban-style stages, stage-change tracking, and stuck-candidate SLA alerts on your Today view.
Google Calendar + Meet invites, RSVP tracking, per-role interview plans, structured feedback forms.
A debrief view and a one-click candidate packet PDF for hiring committees.
Opt-in live sourcing plus an X-ray query builder for LinkedIn, GitHub and resume search.
Per-manager and per-role metrics, interview round stats, and a Looker Studio data tab for live dashboards.
Evaluate candidates with identifying details hidden — fairer first reads, by design.
Admin, Recruiter, Hiring Manager, Interviewer — enforced server-side, with a full audit log.
Automated tracker backups, plus archiving that keeps big pipelines fast without losing your metrics.
MIT-licensed source. Read every line, edit any behavior, swap the AI provider if you want.
Feature-complete is the baseline. Here’s where free-and-open pulls ahead.
Every candidate, CV and score lives in your Google account — not a vendor’s cloud. No lock-in, ever.
You pay Google cents directly for AI. No per-token markup, no surprise “AI add-on” invoice.
Invite every recruiter, hiring manager and interviewer. Unlimited seats, always — because it’s yours.
“Who are my strongest backend candidates?” — agentic search that most ATSs simply don’t have.
Every score shows its reasons across 37 dimensions, with a one-click bias-masked screening mode. Not a black box.
A public careers page plus CV-by-email that reads résumés on arrival and files candidates for you.
Read every line, extend anything, trust what runs. MIT-licensed — a gift to the TA community.
Four things AgentATS does in seconds that used to eat your afternoons. All data below is fictional sample data — animated, not screenshots.
Drop in a CV — Gemini reads it and fills every column for you.
Every candidate scored against the role — then ranked, tiered and ready to review.
Not a black box — every score breaks down into rubric categories you can read and tune.
No filters, no exports. Just ask — and get a ranked answer with reasons.
Fully self-provisioning — it creates its own spreadsheet, folders and tabs. You never touch an ID.
Go to script.google.com and click New project. Free with any Google account.
Copy the six files from the GitHub repo into your project.
One Script Property: GEMINI_KEY, from Google AI Studio. It never lives in code.
One function builds everything and makes you Admin. Deploy as a web app — that URL is your ATS.
~2 minutes. No servers. No bill.
Against the tools it replaces for small and mid-size teams. They're excellent products — they're just built for a different buyer.
| Criterion | AgentATS | Leading paid ATS platforms |
|---|---|---|
| Price | Free forever + cents of AI on your own key | ~$300–$1,000+/mo for small teams |
| Own your data | 100% — the database is a Sheet you can open | Vendor cloud; exports on request |
| AI scoring | Included — a 37-category rubric you can read & tune | Paid add-ons or opaque scoring |
| Open source | Yes, MIT — edit anything | No |
| Setup time | ~2 minutes, self-provisioning | Days to weeks, sales-led onboarding |
If you're a 2,000-person company with a compliance team, buy an enterprise ATS — genuinely. If you're a founder, an agency, or a TA team of one to fifteen, this was built for you.
You're evaluating a hiring system, so here's the full picture up front. Most of these are the deliberate tradeoff of a free tool that runs entirely inside your own Google account.
Apps Script has execution caps and daily quotas (API calls, ~100–1,500 emails/day). Very heavy bulk operations can bump them.
Parsing or scoring a CV isn't instant — expect a few seconds per action. Assistive, not real-time. And AI can misjudge; it assists, it doesn't replace human review.
The database is a Google Sheet: great for hundreds to low-thousands of active candidates, not enterprise volumes. A built-in archiver helps.
Software is free; AI usage is pay-as-you-go on your own key (cents, not seats). Use a paid tier for real candidate data — free tiers may train on inputs.
No SOC 2 / ISO certifications, no third-party pen test. Security is app-level and depends on your Google account hygiene. Regulated buyers need more.
One company per install, no SSO/SCIM, and no built-in LinkedIn/Indeed integrations — intake is your careers page, CV-by-email, or manual.
In one line: the most capable ATS you can run for free inside your own Google account, with real AI — built for startups and lean TA teams, not regulated enterprises. Community open-source means no SLA: you self-host, you self-support, and you own everything.
Star it, fork it, or have it running before your next intake meeting.
★ Get AgentATS on GitHub