Built for Indian data: Aadhaar, PAN, ABHA, Hindi and Hinglish

Share sensitive data, and prove what was removed.

huereo turns a sensitive data export into a copy prepared for one stated purpose, with what was removed and checked on record.

huereo runs inside your own network in India, offline, as one container image. It finds Aadhaar, PAN, ABHA and names in tables, documents and scans, changes them according to a written Privacy Contract, checks the result a second time with an independent verifier, and writes a Passport that records what was done and why.

claims_q3.parquet · row 18,204
namePriya Sharma
aadhaar4821 7730 9926
dob14/03/1987
pin_code411038
phone+91 98220 41177
ifscHDFC0001234
icd10E11.9
claim_inr48,250
clinical_note

Pt Priya S. seen at City Care Hospital on 02/08/2026. Husband Rahul called from 98220 41177 about the claim.

One row as it sits in your warehouse. Nothing checked yet.

A synthetic record, to show the workflow. Not a benchmark.

  1. 01

    Your servers, no network

    One container image. No data, and no information about your data, leaves the machine.

  2. 02

    Tables, documents, scans

    CSV to SQL dumps, Word, email, PDF and scanned images, each released in its own format.

  3. 03

    Data that still works for AI and analytics

    Replace names, phones, Aadhaar and PAN with realistic stand-ins of the same kind and shape, the same for the same person, drawn from ranges no one is issued.

  4. 04

    English, Hinglish and 9 Indian languages

    Hindi, Marathi, Bengali, Telugu, Tamil, Gujarati, Kannada, Malayalam and Punjabi.

  5. 05

    Approved or blocked, never in between

    Only a release that passes every check is written to the shareable folder.

01 Who it is for

Who it is for. India first.

  • Banks and NBFCs
  • Insurers
  • Hospitals and healthcare providers
  • Employers and payroll processors
  • Education institutions
  • Telecom operators
  • Public-sector bodies

Typical shares it prepares

  • A loan book sent to a collections or analytics vendor
  • Patient records sent to billing or research partners
  • Payroll sent to a processor
  • Scanned KYC documents and customer emails sent to an outsourced operations team
  • Records sent to an auditor

02 The problem it solves

Today the shared copy is cleaned by hand.

Someone deletes columns by hand, misses the PAN in a notes field or the phone number inside a scanned letter, and nobody can show afterwards what was shared or what was removed. Under the Digital Personal Data Protection Act, 2023, that risk sits with the data fiduciary.

huereo replaces the manual step with a release that is planned from a contract, executed the same way every time, verified independently and recorded.

03 How a release works

The same eight steps, every release. Every step recorded.

How it works in detail

Step 1 of 8

Discover and profile

Every supported file is read in bounded memory: each sheet of a workbook, each page of a PDF, each part of an email. A file whose name holds personal data (for example PAN_ABCPS1234K.png) is released under a neutral name such as renamed-0001.png.

Produces

  • A profile of every file
  • Neutral file names

04 India first

Built for Indian data.

Identifiers matched with their official structure and checksum, such as Verhoeff for Aadhaar. Names, places and organisations in English, Hinglish and nine Indian languages. A DPDP law pack: a technical mapping of the Act and Rules, not legal advice.

Indian data in detail

  • Aadhaar
  • Aadhaar VID
  • Masked Aadhaar
  • Aadhaar Enrolment ID
  • PAN
  • UPI ID
  • Bank account
  • GSTIN
  • TAN
  • CIN
  • DIN
  • Udyam

Names, places and organisations

  • English
  • Hinglish
  • Hindi
  • Marathi
  • Bengali
  • Gujarati
  • Punjabi
  • Tamil
  • Telugu
  • Kannada
  • Malayalam

05 Working with us

From first contact to production in five steps.

Pilots and licences

  1. Step 1

    Request a pilot

    On this website: sector, what you share and with whom, data types, rough monthly volume, where it would run, timeline. The form never asks for data.

  2. Step 2

    Demo, 30 minutes

    We release a sample loan book, a scanned KYC document and an email live, and walk through the Passport.

  3. Step 3

    Assessment on your machine

    You run huereo assess on one of your own files and get a one-page report of what is sensitive and what to do, plus a drafted contract.

  4. Step 4

    Pilot, 4 to 6 weeks

    We help install the image, write the contract with your privacy team and DPO, and release one real data share. We run huereo review with you to measure detection on your own data.

  5. Step 5

    Annual licence

    One licence file per deployment, sized by volume. Updates arrive as signed files; support works from the support bundle, which holds no data values.

Your data never comes to us at any of them.

06 Questions

Asked in every first call.

Does any data or metadata reach Supermerco?

No. The engine runs offline in your environment and blocks outbound connections. Support works from a bundle that holds no data values, and review scores hold counts only, and the console’s Overview shows counts, names and statuses, never a data value.

What happens when huereo is unsure?

It stops. A field it could not classify, a language it does not cover, a legal decision or a low-confidence OCR page makes the release UNRESOLVED or is reported for review. Nothing unscanned is released as clean. A file with a header and no rows stops with a reason, as an input problem.

Can the recipient link records across files?

Within one release, if the contract allows it: keyed tokens are the same across its files. Across releases only when the contract allows cross-release linking.

Can we see what was done to each field?

Yes. The plan lists every action with its reasons, and the Passport summarises them in plain language for the DPO.

What if a released file is changed afterwards?

huereo verify detects it and names the file.

How do updates reach an offline installation?

As a signed file. You validate and import it, activate it when ready, and can roll back with one command. A file not signed by us is refused.

Does it work with scanned documents?

Yes. Scans and scanned PDF pages are read with OCR in English and Hindi, personal words are painted out of the images, and the released pages are read back by OCR to check.

Which law does it follow?

The contract names it. The DPDP pack maps the Act’s requirements to checks; it is a technical mapping, not legal advice, and the legal decisions stay with your DPO.

Bring one data share you need to send.

Run huereo on one of your own files on a one-hour call, or start a 30-day pilot on your server (₹9,900, credited in full to your plan). Or write to hello@supermerco.com.