Keep fake candidates out of your Greenhouse pipeline.
TalentShield checks every Greenhouse applicant the moment they apply and writes the result onto the candidate record. A badge, a score and the reasons, in plain language. Your recruiters decide.
- The file author in the metadata does not match the candidate's name.Found: author "Admin-PC7", candidate "Jordan P."
- The LinkedIn link in the CV does not open a profile.Found: linkedin.com/in/jordanp-data returns no profile
- Phone number is a VoIP line, not a mobile.Found: number range assigned to an internet calling service
applications flagged by TalentShield and confirmed as fraud by our recruiters, in one month of live hiring
1,768 applications checkedcandidate profiles worldwide will be fake by 2028
Gartner, 2025US companies hired North Korean IT workers through one US laptop farm
US Department of Justice, 2025Catch fake candidates in Greenhouse at the top of your funnel.
Three things land in Greenhouse. Nothing else changes.
A tag with the badge
Looks clear, Worth a look, Needs review or Strong signals. Use it to decide where to look first.
A note with the reasons
The score and the top signals, each in one plain sentence with what we found. A recruiter understands it in ten seconds without opening another tool.
A link to the full report
Every signal with its evidence, decision history and a PDF for your audit trail. One click away, only when someone needs it.
In Greenhouse the result lands as a tag and a note on the candidate, with a link to the full report. Nothing is deleted, moved or rejected on your side.
Greenhouse checks the fit.
TalentShield checks the application.
Greenhouse is built to answer "does this person match the job?". Fake applications are built to pass exactly that test. TalentShield asks a different question.
Already have fraud detection in Greenhouse?
Greenhouse ships its own fraud detection and an identity check partner. Those catch obvious spam and verify an ID at a chosen stage. TalentShield looks at the application itself: file metadata, hidden text, dates that do not add up, contact data, a LinkedIn that does not match the CV. Different checks, and the reasons are written out for the recruiter. Both can run at once. On our benchmark, TalentShield catches 98.7% of fake applications at the top of the funnel.*
*TalentShield internal benchmark, detection engine v0.9, October 2026. Catch rate: share of confirmed fake applications flagged for review. Results on your pipeline may differ.
Three steps to your first checked applicant.
Connect Greenhouse
We connect through the Greenhouse API with access you approve. No plugin for recruiters to install, no change to your process.
Pick the jobs
Switch checks on for the roles that attract fraud: remote, technical, high volume. Leave the rest alone. Old jobs can be checked on request.
Review together
A week later we go through the results with you and tune what runs on your pipeline: which checks, which enrichment sources, how the badge is written back.
See it on your own Greenhouse jobs.
Leave your email. We come back within one business day to book 30 minutes, on your own applicants if you want.
- Pilot on one or two live Greenhouse jobs
- Priced by application volume, not by seat
- Compliance Summary for your security team from day one
Thanks, we've got it.
One of us will reply within one business day. In the meantime, feel free to email contact@talentshield.app.
Greenhouse and TalentShield: common questions
Yes. TalentShield connects to Greenhouse through its API. Every new application is checked as it arrives, and the result lands in Greenhouse as a tag and a note on the candidate, with a link to the full report.
TalentShield connects to Greenhouse through its API and checks every new application against more than 70 signals in four areas: the file itself, what the CV says, contact details and identity, and cross-checks outside the CV. The result, a badge and an Application Integrity Score with the reasons, is written back to the candidate record in Greenhouse.
Start with the file, not the person. Check who created the resume file and when, select all the text to reveal anything hidden in white, and check whether the phone number is a mobile or a VoIP line. Our guide walks through what to check at every stage of remote hiring: How to spot fake candidates in remote hiring.
Greenhouse ships its own fraud detection and an identity check partner. Those catch obvious spam and verify an ID at a chosen stage. TalentShield looks at the application itself: file metadata, hidden text, dates that do not add up, contact data, a LinkedIn that does not match the CV. Different checks, and the reasons are written out for the recruiter. Both can run at once.
No. It adds information to the candidate record for a recruiter to review. There is no automatic rejection, ranking out or filtering, and every recorded verdict is a person's.
Greenhouse stays the source of truth. TalentShield reads the CV and application data to run the checks, keeps only what the report needs and links back to the record. It's stored in the EU under a Data Processing Agreement. AI runs in Microsoft's EU Data Zone and never trains on your data.
Verification runs on your first live job within days of signing, with API access you approve. Pilots start with one or two jobs, a weekly review call, and configuration tuned to your pipeline.
A monthly subscription based on the number of applications checked, not on seats or open jobs. Going over the threshold does not pause the checks. Pilots are month to month, with a founder discount for the first customers on Greenhouse.
Technical details: how TalentShield works in Greenhouse
About Greenhouse
Structured hiring platform used by scale-ups and enterprises worldwide. The most common ATS among tech companies. TalentShield adds the layer Greenhouse was never designed to provide: a check whether the application is consistent with itself and with reality, next to the check whether it matches the job.
How it works in Greenhouse
TalentShield connects through the Greenhouse API. When a new applicant lands on a job with verification switched on, we pull the resume (CV) file and the application data, run the checks, and write the result back to the candidate record. In Greenhouse that is a tag and a note on the candidate, with a link to the full report. Nothing is deleted, moved or rejected on your side; the record simply gets more information on it.
What a recruiter sees
- On the candidate record in Greenhouse: a plain-language badge (for example "Needs review") and the Application Integrity Score.
- One click away: the full report with every signal, what we found, why it matters and how to verify it yourself.
- Across the job: one score per applicant, so recruiters can sort the whole pipeline and start with the strongest signals.
- Never: a "fake" label, an automatic rejection or a filtered-out candidate. Every verdict in the system is a person's.
Switching it on per job
Verification is enabled per job, not per account. Pick the roles where it matters most (remote engineering, roles with access to code, money or customer data, high-volume inbound) and leave the rest as they are. New applicants on those jobs are checked as they arrive; existing applicants can be checked on request. When the signal library updates, you can re-verify a job with one click.
Read more: what candidate fraud detection covers, how the checks work, how we handle compliance in the US and EU, and what a fake candidate actually is.
