Guides · carbon.talent
How to spot fake candidates and verify remote data hires
Last updated 18 Oct 2026
Remote data engineering roles attract templated CVs, proxy interviewers and AI-coached answers. Before you make an offer, check four things:
- they are who they say they are (an ID check, and the same face on every call);
- they can explain their own work in detail;
- they can handle real data scenarios;
- how they actually like to work.
Vetted carbon.talent profiles bring that evidence with the application: a vetting report, a working-style profile and a short video intro. On Team and Scale you can add an ID check.
Warning signs in data applications
- Several CVs with near-identical bullet points. Sometimes the industry doesn't match the employer.
- Every tool in the stack is listed, but no outcome is specific.
- Vague about location, notice period or when a "current contract" ends.
- The camera is "broken" on every call, or the person looks different from their photo.
- Long pauses, then fluent textbook answers. They can't go deeper on their own project.
- They refuse any live or screen-shared work.
A verification checklist, in order
- Before the first call. Read for specific outcomes they owned. Pick one CV bullet and ask a written question about it.
- Video intro. Ask for a short, unedited video explaining a recent project and one decision they'd change. Fraudulent applicants tend to drop out here, and real ones show how they communicate.
- Identity. For finalists, check a government ID with a liveness step. After that, make sure the same person is on every call.
- Scenario interview. Use your own anonymised problem and ask "why" three times. Proxies and scripted answers break down on follow-ups. See data engineer interview questions.
- Short live practical. Screen-shared, with normal tools allowed. Discuss it as they go.
- Working style. A behavioural or work-style profile tells you what to ask about autonomy, communication and handling pressure. Use it to guide the conversation, not to pass or fail people.
- References. Reach them through a company channel you find yourself, not only through contact details the candidate gives you.
- Before access. Don't give access to production data until every check is complete.
What each check catches
| Check | Catches | Doesn't catch |
|---|---|---|
| ID check | A fake or borrowed identity | Weak skills |
| Video intro | A mismatched person, poor communication | Coached answers on their own |
| Scenario questions | Shallow experience, scripted answers | Identity fraud |
| Working-style profile | Fit with how your team works | Skills or identity |
| References | Invented work history | A real person who's a poor fit |
Checks on carbon.talent
- Vetting report from Carbon's real-world data experience test, on every plan. It uses scenario questions on real data judgement. It is not a coding test.
- Working-style profile. Carbon's behavioural assessment produces a named working-style profile.
- Short video intro.
- ID checks on Team and Scale.
- Test a candidate you found. If someone applied through another channel, send them the checks before you make an offer.
Questions companies ask
- How do I know a remote candidate is real?
- Check their ID against a government document with a liveness step, and make sure the same person appears on every call. Ask detailed follow-ups about their own projects, and contact references through channels you find yourself.
- What is a proxy interview?
- Someone other than the applicant takes the interview, or feeds them answers in real time. Detailed follow-ups, screen-shared work and an ID check with the same face on every call make it much harder.
- What does vetted in advance mean?
- Someone checked the candidate before you see them. Ask exactly what was checked: identity, skills, judgement, references. "Vetted" means different things on different platforms.
- Are one-way video interviews useful for data roles?
- A short video intro shows how someone explains their work, and it helps you confirm identity later. It doesn't replace a live technical conversation.
- What is a working-style or behavioural assessment?
- A structured set of work scenarios that shows how someone tends to work: autonomy, communication, handling pressure. Use it to shape your interview questions, not to pass or fail people.