AI lease abstraction reads every lease, deed of variation and side letter in a data room and builds the schedule your team would build by hand, with each value linked to the clause it came from.
All Real Estate & PropTech →
A lease is rarely one document. The original lease is varied by a deed, extended by a reversionary lease, qualified by a side letter on the break and adjusted by a rent review memorandum. The current position is the result of reading them in order. Most abstraction errors, by people or by software, come from reading one document and missing the one that changed it.
So the system is built around the amendment chain. It identifies which documents belong to which demise, orders them by date, and abstracts the current position with a citation to the document that set each term. Where two documents disagree, or a referenced document is missing from the data room, it says so instead of choosing.
scanned and native PDFs, Word, spreadsheets
lease, deeds, side letters, memoranda
by date and by what each document varies
structured fields, clause citation on each
against the vendor rent roll
conflicts, gaps, low-confidence values
Before scoping, we abstract a sample of your own leases that your team has already abstracted, and compare field by field. That tells both of us the accuracy on your documents, including poor scans and older leases, rather than on a vendor demo set.
In production, every value carries its citation, so a reviewer checks by clicking rather than searching. Fields that drive money (rent, review dates, breaks, expiry) are always shown for confirmation, however confident the extraction.
It is the use of language models to read leases and related documents and produce a structured summary of the terms that matter: parties, rent, review dates, breaks, expiry, repair and alienation. Done properly, every value links to the clause it came from and later amendments are applied in order, so the abstract reflects the current position rather than the original lease.
It depends on your documents, which is why we measure it on a sample of your own leases before scoping. Scan quality, handwritten amendments and how consistently the leases are drafted all matter. The design assumes errors are possible: citations make checking fast, conflicts are never resolved silently, and money-critical fields are always confirmed by a person.
Yes. Scanned documents are read with OCR and vision models, and side letters, deeds of variation and memoranda are grouped with the lease they modify. Where a document references another that is not in the data room, the gap is reported rather than ignored.
In whatever your team uses: an Excel schedule for a transaction, or direct import into Yardi, MRI or Argus Enterprise for an owned portfolio. We map our schema to yours so the abstract does not need retyping.
What we built, what stage it is at, and a link to the live thing wherever there is one.
yourOS is an operating system with an AI at its heart. You talk to it like a person, out loud or typed. It makes your calls, plans your days and works on your projects with a crew of agents, on your own private server. Our own product, built from the first commit; the waitlist is open.
Website, backend and app assets for Sonora, a free AI sound-therapy app on iOS and Android published by Cognitive AI.
Brand, website, prototype, apps on five platforms, the AI backend and the algorithm-intelligence pipeline for Podswap, a social growth engine that runs without a human in the loop.
In closed beta: Roxy, a private AI assistant with permanent memory running on infrastructure the user owns. Brand, site, product, backend and conversational design.
Live at coherence.com: brand identity generated with AI agents we directed, the motion system and website for the group behind Sonora, AboutAI, WakeUp and MudShadows. Agent work remains in build.
Live at domainbooth.com for DomainBooth, a premium domain brokerage: a new animated mark, every page redesigned and rebuilt with designed mobile layouts, and a custom CMS for its sales record, featured names, press and enquiries.
Give us ten leases your team has already abstracted. We will compare our output field by field.
30-minute technical call · no deck