Leases, comps, and portfolios, scoped by entity.
Commercial real estate data lives in leases, rent rolls, appraisals, and loan documents spread across property-level entities, funds, and joint ventures. Data.co lets teams query it with private AI while each entity's information stays with the people entitled to it.
Every property is its own set of obligations.
A single sponsor may manage properties owned by different funds, separate accounts, and joint venture partners, each with its own confidentiality terms and reporting rights. Tenant financial statements are often provided under lease covenants that restrict their use. Lender reporting brings its own rules.
Data.co tags documents by property, owning entity, and investor at ingest. Asset managers see their assignments, JV partners see their properties, and acquisitions teams working a competing bid are walled from the incumbent owner's data. Everything runs inside your environment.
Practical work that adds up.
Lease abstraction
Extract rent schedules, options, CAM terms, and co-tenancy clauses into structured fields, each linked to the clause it came from.
Rent roll Q&A
Ask which leases expire in the next eighteen months across a fund, scoped to the properties you manage.
Tenant credit review
Summarize tenant financials for credit decisions, with access limited to the asset team and credit committee.
Investor reporting
Draft quarterly letters from property data for each investor, without exposing other investors' holdings.
Acquisition diligence
Review a target's estoppels, leases, and title documents in a deal-scoped workspace.
From PDF to structured record, with review.
- 1
Ingest
Leases and amendments arrive from the document system, tagged with property and entity.
- 2
Extract
A local model extracts defined fields and records the source page for each value.
- 3
Review
An analyst confirms or corrects fields. Low-confidence extractions are flagged first.
- 4
Publish
Approved abstracts update the lease database, with the change and approver written to the audit log.
Policies real estate teams use.
- Property and owning-entity tags on every document
- JV partner access limited to jointly held assets
- Tenant financials restricted to asset and credit teams
- Acquisition workspaces sealed from asset management
- Export of rent rolls requires approval
- Lender reporting packages generated from scoped data only
Where real estate data comes from, and what restricts it.
| Source | Typical constraint | Default scope |
|---|---|---|
| Leases and amendments | Confidentiality clauses, tenant consent terms | Asset team for the property |
| Rent rolls and property accounting | Owner and investor reporting rights | Owning entity's asset and accounting staff |
| Tenant financial statements | Lease covenants limiting use and disclosure | Asset and credit teams |
| Loan documents and lender reports | Loan agreement reporting and confidentiality terms | Capital markets and asset management |
| Appraisals and broker opinions | Engagement letter restrictions on reliance | Named deal or asset team |
| Licensed market data | Provider license terms | Licensed users only |
Joint venture partners see joint assets.
Properties carry an owning-entity attribute; partner identities carry the entities they invest through. An illustrative rule your team would adapt to its own structure.
id: pol_cre_jv_001
description: JV partners read documents only for jointly held properties
subject: { tags: [jv-partner] }
resource: { collection: "properties.*" }
action: [read, search]
when:
resource.attr.owning_entity: { in: { ref: subject.attr.entities } }
effect: allow
explain: "Partners can access properties held through their joint ventures."Illustrative example · final API surface documented in /docs
Sized for a sponsor, not a hyperscaler.
Most real estate workloads are document-heavy and bursty: lease abstraction runs in batches after an acquisition, and Q&A traffic follows reporting cycles. A private cloud tenancy or a single on-prem GPU server is usually enough to start, and the infrastructure assessment estimates throughput from your lease count and page volume before you commit to hardware.
When evaluating any platform, ask how it represents ownership. If access is managed per folder rather than per entity, a property sale or a new joint venture means re-permissioning by hand. Ask whether an acquisition team can be sealed from asset management during a competitive bid, and whether you can show an investor exactly which identities accessed their reports.
Questions buyers ask
Can JV partners use the system directly?
They can be given identities whose policies limit them to jointly held properties. Many sponsors start with internal use and add partner access later.
How accurate is lease abstraction?
Accuracy varies with document quality and lease complexity, which is why every extracted value links to its source page and goes through human review before it is published.
Do we need to move documents into a new system?
No. Data.co connects to your existing document management, property management, and accounting systems and indexes from there.
Can we use market data alongside our own?
Yes. Licensed third-party datasets can be connected and governed like internal data, subject to each provider's license terms.
What happens when a property is sold?
Its documents keep their entity tags. Policies can move the property to an archived scope with read access limited to accounting and legal, and retention rules decide when derived abstracts and embeddings are removed.
Can lease abstracts feed our property management system?
Approved abstracts can be published to downstream systems through a connector, with the write action governed by policy and usually requiring an approver.
See it against your own use case.
A 30-minute technical walkthrough, policy simulation on your scenarios, and a deployment recommendation for on-prem, private cloud, or hybrid.