AI made it free to claim a skill. So a claim stopped meaning anything.
Resumes are written by AI. Take-homes are answered by AI. Portfolios are generated by AI. When the cost of asserting a skill falls to zero, the assertion carries no information — and every hiring system in the country is still built on top of assertions.
Every hiring system in India runs on unverified input.
A candidate types a skill into a profile. The platform displays it. A recruiter reads it, a matching engine ranks it, an interview panel assumes it. Nothing between the claim and the decision ever checked whether it was true.
Volume was never the problem.
The industry's response has been more tests, more screening, more rounds. Applications per role have doubled since 2022, and roles have become harder to fill, not easier.
Effort used to be the signal.
Writing a convincing resume took effort, and effort was itself a signal. That signal is gone — what used to take an afternoon now takes a prompt.
The missing piece is trust, not data.
More data about candidates was never the fix. The missing piece was always knowing which data to believe.
The question is no longer whether someone claims a skill. It is what stands behind the claim.
Not all skills are equal. Proof of work outranks proof of exposure.
All evidence is not equal, and pretending otherwise is how trust frameworks become decoration. The ladder is ordered by how directly the evidence touches work that was actually done.
A reporting manager watched the work happen. A public repository or portfolio is the work, in the open, timestamped. A proctored assessment simulates the work under observation. A course certificate proves someone was taught, not that they can do it. An experience letter proves employment, not capability. A self-declaration proves intent and nothing more.
Every one of those is real evidence. They are not interchangeable, and averaging them into a single score destroys the only information that mattered. So Record doesn't produce a score. Each verified skill carries a label naming what stands behind it, who provided it, and how it was obtained. A recruiter can see the difference between a skill vouched for by a reporting manager and a skill backed by a training certificate — because those are different facts.
Source is a choice. Trust level
is a fact.
Seven levels of proof.
Source is a choice. Trust level is a fact. Seven levels for white collar work, two levels for blue and grey, one label per verified skill.
Levels 1–2
A person puts their name on it. Someone who managed or worked directly alongside the candidate confirms the skill from a verified work email. The gap between the two is proximity. Every endorsement carries its verifier's identity, role, relationship and location — visible to the next recruiter. It is not anonymous and it is not free.
01 · Highest
Current/previous employer — reporting manager
02 · Very high
Employer — founder, CXO, HR, team lead
Levels 3–4
The work is visible, but nobody vouched. A public repository, a solved problem set, a shipped portfolio outranks a certificate, because it is output rather than attendance. A proctored assessment sits just below it — generated for that candidate alone, observed throughout.
03 · High
GitHub, HackerRank, LeetCode, Behance
04 · High-medium
Proctored — MCQs, coding & design sandboxes
Levels 5–7
Proximity to the work drops away. A training partner confirms someone was taught. A payslip confirms someone held a role. Neither confirms they can do the thing. All three are routinely mistaken for proof — which is how most hiring decisions in the country are currently made.
05 · Medium
L&D partner — EdTechs, institutions, academies
06 · Medium-low
Experience letters, pay slips — OCR + AI
07 · Low
Self-declared — by the candidate
Two levels of proof.
High-medium
Video verification
Method
Recorded scenario responses in the candidate's language, evaluated for skill and authenticity together
Proves
Skill and identity, with reviewable evidence
LOW
Voice verification
Method
Structured phone conversation where only what is said can be evaluated
Proves
Intent and stated experience
Why the ceiling is Medium
The top of the framework requires a human touchpoint — a former employer who can be identified, reached, and held to what they say. For most blue and grey collar work in India, that person cannot be traced. Employment is informal or contracted through intermediaries; there is often no record the working relationship existed at all.
So the highest available proof for this workforce is different in kind: a recorded video, in the worker's own language, answering scenario questions about the trade, evaluated for skill and authenticity together. That is real evidence — reviewable, attributable to a person, and more than any employer in India currently asks for. It is also the first time pre-employment video proof has been available to this workforce at all.
It sits at Medium because the framework is consistent across every kind of work. The ceiling reflects what can be traced, not what a person is worth. If a route to verifiable employer endorsement for informal work becomes feasible — and we are actively looking for one — the ceiling rises.
The same standard applies to a software engineer and a welder. The difference is what can be traced, not who is being judged.
We describe the proof. You make the decision.
A trust level is not a rating of a candidate. It says nothing about how good someone is at their job — only what kind of evidence stands behind a specific claim about a specific skill.
Record does not decide who is qualified, does not rank people against each other, and does not set a floor on who a recruiter should see. A hiring team knows things about their own role that a framework never will. Our job is to make sure they are deciding with real information rather than assumed information.
The label is visible. What to do with it is yours.
Use it. It isn’t ours to keep.
We built this framework because we needed one and none existed. We are publishing it because a vocabulary for skill proof is more useful to India than a vocabulary owned by one company.
Platforms, staffing companies, assessment providers and background verification firms are welcome to adopt these levels, argue with them, or improve them. If the industry converges on a shared language for what stands behind a skill, verified credentials start to travel – across platforms, across employers, across a working life. A skill proven once should not need proving again at every door.
That is the outcome worth building toward. The infrastructure that produces the proof is our work. The standard that describes it should belong to everyone.
India’s next economic leap will not be built on degrees. It will be built on verified skills.
Record is the infrastructure that produces the proof for 390 million workers across white, blue and grey collar work.