AI & Data
Structured vs Unstructured Financial Data
Most private markets data starts life unstructured. Understanding the difference is the first step to fixing it.
Definitions
Structured financial data is organized in a predefined format - rows, columns and defined fields - so software can query and analyze it directly. Unstructured financial data has no fixed data model, such as PDF reports, presentations and narrative text, and must be interpreted before it can be analyzed.
Examples in Private Markets
| Structured | Unstructured |
|---|---|
| Accounting system ledgers | GP quarterly and annual reports (PDF) |
| Capital call data feeds | Investor letters and presentations |
| Normalized portfolio databases | Portfolio company narrative updates |
| Public market price feeds | Scanned documents and images of tables |
Why GP Data Is Unstructured
GP reports are designed for human readers. Tables are formatted for presentation, context sits in headers and footnotes, and every manager uses a different layout. The data is there - but not in a form software can use.
Semi-Structured Data
Tables inside PDFs are sometimes called semi-structured: they have visual structure, but no machine-readable schema. Extracting them reliably needs layout-aware document AI rather than plain text extraction.
Converting Unstructured to Structured Data
1 Source PDF
PDF Q2 2026 GP Quarterly Report · p.14
Revenue £125.4m
EBITDA £31.2m
EV £420m
2 Zanran structured data
| Metric | Value | Source |
|---|---|---|
| Revenue | £125.4m | Q2 2026 GP Report |
| EBITDA | £31.2m | Q2 2026 GP Report |
| EV | £420m | Q2 2026 GP Report |
3 Portfolio intelligence
EBITDA margin 24.9%
EV / EBITDA 13.5x
Traceable page & cell
Illustrative figures. Every value in Zanran links back to its location in the original report.
Zanran uses computer vision and machine learning to read report pages, extract values with their context and map them into a standardized model. AI financial data extraction.
FAQ
Frequently Asked Questions
Straight answers to the questions investment teams ask most. Still curious? Talk to us.
What is the difference between structured and unstructured financial data?
Structured financial data sits in predefined fields that software can query directly; unstructured financial data, such as PDF reports and narrative text, has no fixed data model and must be interpreted and extracted before analysis.
Is a PDF table structured data?
Not for software. A table in a PDF has visual structure but no machine-readable schema, so it is usually described as semi-structured and needs layout-aware extraction.
See Zanran in Action
Upload a set of GP reports and see asset-level data extracted, normalized and traceable to source - in minutes.