Privacy Policy
This policy describes what the platform and its browser extension collect, why, where it is kept, and what is never done with it. It covers the web application, the API behind it, and the Nirnaya browser extension.
Last updated 11 September 2026
What this product does
Nirnaya is an investigation tool for Indian public records. An investigator enters a person or a company, and the platform searches public court, tribunal and registry sources — the Supreme Court, High Courts, district judiciary, the NCLT, the DRT, CERSAI and the MCA — then assembles the matching public filings into a single profile with citations back to the source.
Everything the platform gathers is a record those bodies publish. The product does not create facts about people; it locates, links and organises records that are already public, and records honestly where it has not looked.
Information collected from account holders
Accounts are issued to firms, not opened by the public. For a person using the platform, the following is collected:
- Account details — name, work email address, the organisation the account belongs to, and the role assigned within it.
- Authentication data — a password, stored only as a salted hash, and a session cookie that identifies the signed-in session. The cookie is required for the product to function and is not used for advertising or cross-site tracking.
- Work product — the investigations created, the subjects entered, the searches run, and the documents and reports produced.
- Operational logs — timestamps, request paths and error traces needed to keep the service running and to diagnose failures. These may incidentally include an IP address.
Information about people who are not users
An investigation necessarily concerns people who have no account and no relationship with the platform. A search returns court filings, tribunal orders and registry records that name parties, directors, borrowers, guarantors and their counsel. Those names, and the details printed alongside them in the public record, are stored as part of the investigation they were found in.
Names are stored exactly as the source printed them. Deciding that two records describe the same person happens in a separate, reviewable step, and never overwrites the original text — a wrong merge is worse than a duplicate, because a duplicate is visible and a merge destroys a name.
A person who appears in an investigation may ask what is held about them and ask for a correction. Because the underlying records are public filings held by courts and registries, a correction here changes what this platform shows; it does not and cannot change the public record itself, which only the issuing body can amend.
The browser extension
The extension opens in a side panel and drives the public search forms on court, tribunal and registry websites, capturing each result as a document and filing it into an investigation. It runs only on the sites listed in its manifest, and only while a collection run is active.
Each permission it requests exists for one reason:
- Side panel — to present the extension's own interface.
- Tabs — to open and drive the portal tab a collection run uses, and to tell a loaded search form from a stale results page.
- Storage — to keep a resume cursor, so an interrupted run costs one search rather than the whole run. It holds which searches are complete, not the contents of the pages.
- Debugger — to print a rendered portal page to PDF with its text intact. Several sources publish case details as a web page and nothing else, so the print is the citable record. This is why the browser displays a debugging notice during a run. It is used for printing pages on the listed portals and for nothing else.
- Site access — limited to the specific court, tribunal and registry origins the collectors use, plus the product's own website and API. No permission is requested for payment pages: where a registry charges for a search, the payment happens in a window this extension cannot read.
The extension requests no download permission and writes nothing to your computer's file system. Collected documents go directly to the account's investigation on the server.
The extension does not read pages you visit outside the listed sources, does not collect browsing history, and does not record keystrokes, form entries or credentials on any site.
How information is used
Collected information is used only to:
- run searches and assemble the resulting records into an investigation,
- link records that refer to the same case, person or company, and record where a link could not be made,
- produce reports and exports for the account holder,
- authenticate sessions and enforce which organisation may see which investigation,
- keep the service running, diagnose faults and prevent abuse.
Investigation content may be sent to a third-party language model provider to extract structured fields from a collected document and to answer questions about an investigation's own records. Content sent this way is used to serve that request. Counting and totals are computed by database queries, not by a model, so figures reported are read from the stored records rather than estimated.
What is never done
- Personal information is not sold, rented or traded.
- Data is not shared with third parties for their own marketing, advertising or profiling.
- Data collected through the extension is not used for advertising, analytics resale, or building any profile beyond the investigation it was collected for.
- Investigations are scoped to the organisation that created them; one client's work is not visible to another.
- No credit-scoring, employment-screening or automated eligibility decision is made by this product.
Sharing and disclosure
Information is disclosed only to service providers that operate the platform on our behalf — hosting, database, storage and the language model provider described above — each acting on instructions and only to deliver the service. Information may also be disclosed where the law requires it, or to establish or defend a legal claim.
Retention
Investigation records, collected documents and the reports built from them are retained for as long as the account holder's organisation keeps the investigation, and are deleted on request. Operational logs are kept only as long as they are useful for diagnosis and security. When an account is closed, its investigations are deleted or returned to the organisation that owns them, at that organisation's direction.
Security
Traffic is encrypted in transit. Passwords are stored only as salted hashes. Access to an investigation is restricted to the organisation it belongs to and the roles within it. No system is perfectly secure, and no absolute guarantee is offered — but a breach affecting personal information will be reported to the affected account holders without undue delay.
Your choices
An account holder may access and correct their own account details from the settings screen, export the investigations they have access to, and ask for an investigation or an account to be deleted. There is no self-service password reset by design: accounts are administered by the firm that holds them, and a reset is requested from that firm's administrator.
The extension can be removed from the browser at any time from the browser's extensions page. Removing it stops all collection; it does not delete records already filed into an investigation, which are deleted through the platform.
Children
This is a professional tool. It is not directed at children and accounts are not knowingly issued to anyone under 18.
Changes to this policy
This policy may be updated as the product changes. The date at the top of this page records when the text was last revised, and material changes will be notified to account holders before they take effect.
Contact
Questions about this policy, requests for access, correction or deletion, and any privacy complaint should be sent to the administrator of the organisation that issued your account, who will escalate to the platform operator. A person named in an investigation who does not hold an account may contact the platform operator through the address published on the listing or website from which this product was obtained.