How We Build Intelligence
Primary-source extraction. Structured normalization. Confidence-scored resolution. Capital graph assembly.
Every record in the system traces back to a verifiable institutional filing or public record. We do not begin with third-party aggregation, broker marketing materials, or inherited assumptions. We begin with source records, then build upward into a structured intelligence layer for private CRE credit.
Four Pillars of Intelligence
Most CRE data products start with aggregated datasets. Ours starts with primary-source extraction and builds through normalization, entity resolution, and graph assembly.
Primary Source Extraction
We build from regulatory filings and public records, not third-party databases. Every material data point is anchored to a verifiable source. No scraped broker listings. No inherited aggregation layers. No opaque sourcing.
The raw institutional record is the starting point for everything we produce.
Cross-Regime Normalization
Institutional debt regimes differ in filing format, reporting cadence, and field structure. We normalize them into a single queryable schema with consistent definitions across records.
One schema. One API. One connected intelligence layer. This removes the stitching burden from the client side and makes the dataset usable across origination, portfolio risk, and model-driven research workflows.
Entity Resolution
Institutional CRE lending often sits behind single-purpose entities that obscure the underlying sponsor, operator, or parent relationship.
Our resolution process maps from borrower LLC to operating entity to parent counterparty, assigning confidence scores at each step. The goal is not to overstate certainty, but to make legal structures analytically usable while preserving provenance and confidence at every link.
Capital Graph Assembly
We connect allocators, lenders, sponsors, borrowers, and properties into a single capital graph.
This creates visibility across the full credit chain, from macro capital allocators to property-level debt exposure. The graph is confidence-aware, not assumption-based, and is designed to surface concentration risk, capital family dependencies, and cross-lender relationships that remain invisible in isolated records or within any one institution’s internal book.
Confidence at Every Level
Every material record carries a confidence designation. Every relationship is scored. The system indicates not only what is known, but how strongly it is supported.
Multi-source verified
Validated through two or more independent high-reliability sources. This is the highest confidence designation in the system.
Single high-reliability source
Supported by a single high-reliability source and retained pending independent corroboration. Strong signal, clearly labeled.
Derived from verified inputs
Computed from verified underlying data using defined logic or validated models. Derived maturity metrics, refinancing pressure indicators, and other calculated fields fall into this class.
Queued for manual review
Insufficiently supported for downstream analytical use without additional validation. Flagged for review before inclusion in higher-confidence workflows.
What We Exclude
What we exclude matters as much as what we include. These boundaries protect provenance, data integrity, and client trust.
No Third-Party Aggregated Data
We do not rely on aggregated property databases or inherited data layers. Records originate from primary regulatory filings and public records.
No Scraped Broker Listings
We do not incorporate brokerage marketing materials, listing data, or scraped deal flow into the core intelligence layer.
No Untraceable Records
If a record cannot be tied back to a verifiable source, it does not enter the system.
No Opaque Sourcing
We do not ask clients to trust black-box record construction. Provenance is part of the product.
Continuous Coverage, Structured Verification
The platform is maintained through continuous source monitoring, rolling refresh cycles, and periodic verification against institutional filing cadences.
This methodology is what makes the platform usable as infrastructure rather than just reference data.
The result is a private credit intelligence layer built for origination, risk, and research teams that need structured signal, not surface-level aggregation.
Primary-source intelligence, not aggregated noise.
Request beta access and a data sample scoped to your geographies, workflows, and credit questions.
Request Access