How to verify an AI engineer’s credentials

Check a credential where it is issued, not where it is advertised. As of 2026, a third-party badge, an auditable track record and production systems you can question tell you more than a percentile a platform grades for itself. Super Cat Technology publishes the same checks for its own team.

Cat Yung
, Founder, Super Cat Technology
Published · Updated
credential.check4 checks
  • Vetting tierbadge awarded by a third partyverified
  • Auditable track recordJob Success Score + contract historyverified
  • Production, not notebooksask what runs todayask
  • Explains in business termsfirst callask
check a credential where it is issued, not where it is advertised
Illustration

What does “vetted” actually mean?

It depends on who did the checking. Across AI-staffing marketplaces it means one of three things:

Kind of claimExampleCan you check it?
Self-graded percentileA platform’s own “top 3%” of applicantsNo: the platform grades its own roster
Speed as proofA promise of fast matchingNo: speed says nothing about skill
Third-party badgeUpwork Expert-Vetted, issued by UpworkPartly: Upwork maintains it, but only Enterprise and Business Plus clients see it

Self-graded claims are not necessarily false. They are just each provider’s own statement, so the first real test of the engineer becomes your project. A third-party badge is stronger because another organisation stands behind it. Check that you can see the badge before you rely on it.

Which four checks should you run?

All four take under ten minutes from public profiles and a short call, with no technical interview:

  1. A verified vetting tier. A badge the platform awards after its own review, not a line on a CV. Check that your account can see it.
  2. A track record you can audit. A Job Success Score and contract history across years, with reviews written by clients.
  3. Production, not notebooks. Ask what runs inside a real company because of them. Notebooks show someone can model; production references show they can ship.
  4. Business-terms communication. Can they explain the trade-off they made, and what it cost or saved, without hiding behind jargon?

How do you check an Upwork Expert-Vetted badge?

Open the engineer’s Upwork profile, then check which badges your account can see. Upwork shows Expert-Vetted only to Enterprise Suite and Business Plus clients; other clients see the highest public badge, usually Top Rated Plus, plus the Job Success Score and contract history.

Upwork says Expert-Vetted is reserved for the top 1% of talent on the platform and screened by its own talent managers (source: Upwork, 18 November 2026). Upwork’s badge-display rules list which badges each client tier sees. Rising Talent, Top Rated and Top Rated Plus are visible to all clients; Expert-Vetted is not (source: Upwork, 18 November 2026).

If you are not on one of those plans, check three public signals. Look at the Job Success Score, the contract history with client reviews, and the Top Rated Plus badge. Treat an Expert-Vetted claim you cannot see as a claim, not a credential. If a provider cannot point you to a public profile with a work history, its vetting is a claim you cannot check.

What does an identity badge prove?

An identity badge proves that a platform checked who the person is. It does not prove that they can do the work. Upwork’s ID Verified badge and LinkedIn’s identity verification sit next to skill signals on the same profile, so read which claim each badge makes.

Upwork’s ID Verified badge means Upwork has seen a government photo ID and confirmed a phone number. In many cases it has also checked the person on a short video call (source: Upwork, 18 November 2026). It says nothing about skill, work quality or the accuracy of the work history.

LinkedIn verifies identity and workplace, and is testing peer confirmation of work and study history. On 23 September 2026 it said more than 115 million members had verified their identity or workplace (source: LinkedIn, 23 September 2026). That is a real signal against impersonation. It is not a technical assessment.

A useful rule: separate the claim from the check. Identity, work history and skill are three claims, and each has a different issuer. Ask which issuer stands behind the one your project depends on.

What should you ask about production work?

  • What is running today because of you, and for how many users or employees?
  • What broke after launch, and how did you find out?
  • How is it evaluated and monitored, and who gets paged?
  • What did you hand over when you left, and where does it live?

What are the red flags?

  • A percentile with nothing behind it you can open and read.
  • A portfolio of notebooks and demos with no system in production.
  • “Experienced with LLMs” with no named deployment or scale.
  • Speed offered as the main proof of quality.

What does a verifiable AI engineer look like?

Hold every candidate to the same standard, including us. Here is how Cat Yung, Super Cat’s founder, scores on the four checks:

  • Vetting tier: Expert-Vetted on Upwork, top 1% platform-wide, visible to Enterprise Suite and Business Plus clients (source: Upwork profile, 18 November 2026).
  • Track record: Top Rated Plus, 100% Job Success Score, visible on the public profile (source: Upwork profile, 18 November 2026).
  • Production: a 24/7 AI voice agent at a 5,000-employee US enterprise and enterprise RAG inside a Big 4 accounting firm, written up on the portfolio.
  • Professional history: on LinkedIn.

What can a credential not tell you?

A badge proves who checked an engineer, not that they fit your problem. Even a verified engineer needs a scoped first piece of work, access to your data and an owner on your side. Treat the credential as the reason to start, and the first few weeks as the real interview.

If you want the four checks applied before the first call, see hire vetted AI engineers.

Frequently asked questions

Is an Upwork Expert-Vetted badge worth checking?

Yes, if your account can see it. Upwork shows Expert-Vetted only to Enterprise Suite and Business Plus clients, and says it is reserved for the top 1% of talent, screened by its own talent managers. On the public profile, use the Job Success Score, contract history and Top Rated Plus alongside named production work.

Is “top 3%” a meaningful credential?

It is the platform’s own claim about its own applicants. Toptal says fewer than 3% of the applications it sees each month are accepted (source: Toptal, 18 November 2026). You cannot check the number independently, so use it as context rather than proof.

How long does it take to verify an AI engineer?

The four checks take under ten minutes from public profiles and a short call: vetting tier, auditable track record, production evidence, and business-terms communication. That verifies the claims. It does not replace a scoped first piece of work, which is where you learn whether the engineer fits your problem.

Do I still need a technical interview?

Not to verify credentials: a third-party badge, a public work history and named production systems answer the credential question. You do still need a scoped first piece of work with your data and your constraints, because a credential proves who checked the engineer, not that they fit your problem. Treat the first few weeks as the real interview.

What is the difference between an identity badge and a skill badge?

An identity badge proves a platform checked who the person is, usually with a photo ID and a live check. A skill badge proves someone assessed their work. Upwork’s ID Verified and LinkedIn’s identity verification are identity checks; Expert-Vetted and Top Rated Plus are skill and delivery signals. Check the one your project depends on.

How do I hire vetted AI engineers for an enterprise?

Ask who did the vetting, then check it yourself. A third-party badge, a public work history and production systems you can question beat a self-graded percentile. As of 18 November 2026, Super Cat Technology’s engagements are led by Cat Yung, Expert-Vetted top 1% on Upwork with a 100% Job Success Score, and the team embeds on a time-and-materials basis.

Sources

  1. Upwork, What is Expert-Vetted status on Upwork? (accessed 18 November 2026)
  2. Upwork, How talent badges display on your Upwork profile (accessed 18 November 2026)
  3. Upwork, How to get the identity verification badge (accessed 18 November 2026)
  4. LinkedIn, Verifications on your LinkedIn profile (23 September 2026)
  5. Toptal, Top 3% (accessed 18 November 2026)
  6. Upwork profile, Cat Yung (18 November 2026)
Hire vetted AI engineers →

Engineers whose vetting a third party audited.

Best freelance data scientist, 2026 →

Platforms, rates and a shortlist.

Toptal alternative for AI →

Marketplaces and specialists, compared.

About Cat Yung →

The credentials above, in full.

Cat Yung

About the author

Cat Yung is the founder of Super Cat Technology, an AI agent engineering team in Hong Kong and London, and works as an AI expert, data scientist and fractional CTO. Over a decade in production AI, NLP and machine learning; Expert-Vetted top 1% on Upwork with a 100% Job Success Score (source: Upwork profile, 18 November 2026).

First-Sprint Data Scientist

A vetted engineer embedded in your team, a shipped deliverable by end of sprint 1.

Pick a time below and you will speak with me directly. Bring the use case, the data and the constraints, and I will give you a straight view on feasibility, effort and the first thing to ship.

Book a call

Cat Yung

Hong Kong & London time zones

Book a call · 30 min with Cat Yung