A joiner is twenty fields somebody has to type
Name, father’s name, Aadhaar, PAN, bank account, address, emergency contact, qualification. Multiply by every joiner, and onboarding becomes a transcription job that nobody has time to do accurately.
Onboard without data entry
Documents Become Records
Upload the Aadhaar, PAN and joining form; Klok drafts the fields and flags what it is unsure about for you to confirm.
The Joiner Fills Their Own Details
Send an invite; they complete their information and upload documents from their phone before day one.
Nothing Forgotten
Documents, bank details, statutory numbers and the first-day checklist tracked in one place rather than one person’s memory.
The transcription is the whole cost
A joiner is not a difficult record. It is twenty straightforward fields — and twenty opportunities to introduce a spelling that fails a PF claim eighteen months later.
The fields already exist, in the documents the joiner is holding. The work is not collecting the information; it is retyping it from an image on a phone into a form on a laptop, once per joiner, accurately, while doing four other things.
Two changes remove it
First, the joiner enters their own details, from an invite, on their own phone. They know how their name is spelled and which of the two dates on the certificate is the right one.
Second, their documents are read rather than transcribed. Aadhaar, PAN, degree certificate and bank details are dropped in together, read on arrival, and drafted into fields the joiner confirms.
What is left for HR is verification — checking that the document is legible, current and the one that was asked for. That is a genuinely different task from typing, and it is the one that actually requires judgement.
What this prevents downstream
A UAN captured correctly at onboarding prevents a duplicate account. A name matching across UAN, Aadhaar and PAN prevents a stalled claim. A date of birth entered from the certificate rather than from memory prevents a correction request that takes months.
None of these are onboarding problems when they happen. They are onboarding problems that surface much later, which is exactly why they are worth removing at the point of entry.
Twenty fields, none of them typed by you
Send the invite
A link to the joiner. No account creation, no training, no laptop needed.
They drop their documents
All of them at once, photographed on their phone.
Each is read
Fields drafted from the documents, with anything uncertain flagged rather than guessed.
You verify
Check and confirm — the judgement task, not the typing task.
Onboard without data entry FAQs
What if the joiner has no smartphone?
HR can complete the same flow on their behalf, dropping in photographs of the documents. The reading still happens; only the data entry moves.
How accurate is the reading?
Good on printed documents, less good on damaged or handwritten ones. It flags what it is unsure about rather than guessing, and every field is confirmed by a person before it reaches the record.
Does the document stay in the system?
Yes, in the document vault, attached to the employee, with its expiry tracked where it has one.
Can we still require a physical file?
You can, and some employers must. This changes what goes into the file, not whether you keep one.
How much does the employee do themselves?
They complete their own details and confirm what the AI read from their documents. HR verifies rather than transcribes.
What if a document is unreadable?
It is flagged rather than guessed at, and a person fills the gap. A confident wrong value is far more expensive than an honest blank.
Try it on your own payroll
Fourteen days, no card, and every record exports back out if you decide against it.