Reading the easements you already hold
Four thousand easements in the file, and the client wants to know which ones allow a second line.
Revelry Ai reads legacy grants into one table, cited to each instrument.
The answer is in the file somewhere
Due diligence answers sit in the grants themselves.
Answering one means a team reading every grant on the system.
A client buys a system, plans an expansion, or wants a new line in an existing corridor. Someone has to read every grant on it. Most are scans. Some are typed, a few are handwritten, and the forms change every decade.
- Width is fifty feet on one tract and "sufficient for the purposes herein" on the next.
- One grant allows additional lines. The one beside it limits the corridor to a single pipeline.
- Assignment language, access rights, abandonment clauses, and surface restrictions change tract to tract.
Your team reads each one, keys the terms into a spreadsheet, and reconciles it against the tract list.
Every grant goes into one table.
- The terms your client asked about get pulled from every instrument: width, rights granted, number of lines, assignability, access, restrictions.
- Each cell cites the instrument and the page where the language sits.
- Grants that say nothing on a term get marked silent, so silence becomes a finding your team can see.
Then ask it.
Which tracts on this segment allow a second line? Which grants limit access to a single road? The answer comes back as a list of tracts with the source language attached.
Built against your questions
The fields get set during setup to the questions your client is asking. A collocation study reads for different terms than an acquisition audit. That configuration is part of the engagement.
What it does not do
It does not give a legal opinion on what a grant permits. Ambiguous language gets flagged with the text attached for your counsel or examiner to decide. Faded or handwritten grants it cannot read with confidence go to a person.
Easement review runs on the same reading as title review.
Instruments get read in your format first. Then the terms your client asked about get pulled into one table your team can query.
Bring a segment of grants.
Pick ten tracts and one question your client asked. Watch it answer with the language attached.