About us
Indian Litigation OS is a research and drafting workspace for people who appear in Indian courts and tribunals, and for the mooting teams who argue as if they did. It exists to shorten the first pass over a file without becoming another thing you have to check.
Legal research tools are good at finding cases. What they are not good at is the distinction a litigator makes without thinking about it: between a case that came up in a search and a case you would put in a list of authorities. Those are different things, and treating them as the same is how a judgment from the right court in the right field ends up cited for a proposition it never decided.
The consequence lands on one person, in public. Opposing counsel reads the judgment; the bench reads the judgment; you relied on a summary. That single failure mode shaped the whole application.
There are two lists and they are never merged. Research candidates are judgments worth reading. Final authorities are judgments you can stand up and rely on. A case reaches the second list only when nine separate conditions are all satisfied — set out in full on the home page — and the load-bearing one is deliberately mechanical rather than a matter of judgment:
The words said to establish the proposition must appear in the judgment, character for character. A passage that was composed, paraphrased, tidied, or stitched together from two places will not be found, and the authority fails the check.
This is why the system will sometimes tell you it found nothing. An empty list with an honest note is a result you can work with. A list padded with cases that merely resemble yours is a liability that reads like research.
Before any case is looked for, the system settles what law actually applies: the forum, the statute, treaty or rules that govern, whose decisions bind that forum, and whether the record is still open. This ordering is not cosmetic. Rank authority without it and you get a confident answer to the wrong question — Indian domestic judgments presented as the applicable law in a matter governed by an international convention, or evidence suggested on a record that closed months ago.
For a moot, that means Indian cases are correctly marked as comparative material and no one is told to go and gather more evidence. For live litigation, it means the opposite, because there the record is open and gathering evidence is the point.
If you hold material that cannot leave your own infrastructure at all, the analysis step is the one to discuss before you upload it.
We do not give legal advice, and this application does not either. It does not decide what to argue, choose between strategies, or tell you your prospects. Every draft is a first draft for you to settle. Every authority on your list is one you should read before you rely on it — the checks exist to make that reading worth your time, not to stand in for it.
We also do not pretend to certainty we do not have. Where the system cannot establish something, it says so in those words, and where it has withheld an answer it tells you why. If you ever find it asserting something it cannot show you the source for, that is a defect, and we would like to hear about it.
This is an early release, in use and improving. The honest limitations today: retrieval is keyword-based rather than semantic, so the validation stage can only judge what the search surfaced; the library of judgments grows as matters are researched rather than arriving complete; and a full first pass over a heavy file takes a few minutes rather than seconds, because the checking is done properly.