Discovery lets you analyze many documents at once by asking the same questions across all of them. Results appear in a structured table, making it easy to compare information and spot patterns.
What is a Discovery?
A discovery is a systematic extraction of information from multiple documents. Think of it as creating a spreadsheet where Libra fills in the data by reading your documents.
| Element | Purpose |
|---|
| Template | Defines the columns (questions) to answer for each document |
| Documents | The files you want to analyze |
| Results | A table with one row per document, showing extracted data |
When to Use Discovery
Discovery is ideal when you need to:
| Use Case | Description |
|---|
| Extract key data | Pull specific information (parties, dates, amounts) from multiple contracts |
| Compare provisions | See how terms vary across a set of documents |
| Identify patterns | Spot inconsistencies or trends in document sets |
| Build summaries | Create overviews of large document collections |
How Discovery Works
- You define columns - the questions you want answered for each document
- You upload documents to analyze
- Libra extracts information from each document for each column
- Results appear in a table with one row per document
The key benefit is scale—you can analyze hundreds of documents with the same questions, getting consistent, comparable results.
Discovery vs. Review
| Discovery | Review |
|---|
| Focus | Data extraction | Detailed analysis |
| Documents | Many | One or few |
| Output | Table/spreadsheet | Structured report |
| Best for | Comparing across documents | Deep dive into one document |
Getting Started
