A senior advocate in Bengaluru once told us that a clean title search is mostly patience. You requisition the encumbrance certificate, wait for the sub-registrar, read three decades of deeds by hand, reconcile the survey number against the RTC, and hope nobody asks you to do it again next week. For a single home loan, that round trip routinely takes two to three weeks — and the bank's borrower has usually found another lender by then.
We built LegiScore to compress that work without removing the lawyer from it. The target was deliberately unglamorous: take the same records a panel advocate reads and return the same conclusions, but in fifteen minutes instead of fifteen days. Here is what that actually involves.
The records were never the bottleneck — reading them was
Karnataka's land records already live online. Bhoomi holds the RTC (Record of Rights, Tenancy and Crops), Kaveri carries registered deeds and encumbrance certificates, and Telangana's Dharani exposes the integrated land record. The data is there. The problem is that it arrives as scanned Kannada and Telugu pages, inconsistent survey-number formats, and EC entries that reference deeds you then have to go and pull separately.
LegiScore ingests all of it and normalises it into a single structured timeline. Vernacular OCR tuned on land-record vocabulary handles the script. A reconciliation layer matches survey and sub-division numbers across portals so a plot recorded as “45/2B” on Bhoomi and “45-2-B” on Kaveri is treated as one property, not two.
Tracing the chain of title, automatically
Once the records are structured, the system walks the chain backwards. Every transfer — sale, gift, partition, inheritance, mortgage and release — becomes a node. LegiScore checks that each link is whole:
- Does every owner in the chain have a registered instrument transferring title to them and from them?
- Do the EC entries reconcile with the recited deeds, or is there a transaction nobody declared?
- Are there subsisting mortgages, court attachments, or lis pendens that survive into the present owner's title?
- Does the extent in the deed match the extent in the revenue record, or has the plot quietly grown?
Anything the system cannot resolve is flagged, not hidden. A break in the chain, an EC gap, or a name mismatch surfaces as an explicit exception with the underlying page attached — so the reviewing advocate spends their time on the three entries that matter, not the ninety-seven that are clean.
A report a credit committee can actually bank on
The output is not a chat transcript. It is a structured title opinion: the chain of title as a timeline, a list of encumbrances with their status, the searches performed with their sources, and a clear marketability conclusion. Every assertion links back to the exact deed or EC entry it rests on. When a credit committee asks “how do we know the 2014 mortgage was released?”, the answer is one click away, not one phone call.
On a typical residential file, that takes LegiScore about fifteen minutes end to end. The advocate then reviews the exceptions and signs. The judgment stays human; the patience becomes a machine's problem.
What changes when due diligence stops being the long pole
When a title search collapses from weeks to minutes, the economics of lending shift. A loan officer can clear a property the same day a borrower walks in. A law firm can take on volume it used to turn away. And the bank stops losing good borrowers to whoever could say yes first. The diligence does not get thinner — it gets faster, more consistent, and fully traceable. That is the whole point.