Create A Box Metadata Extraction Scanner
This guide covers the process of creating a Hydra Scanner to extract metadata from documents stored in Box. As an example, you can create a Hydra Scanner to extract metadata such as employee name and contract date from employment contracts.
Prerequisites: Create a custom Box Skill and map it to Hydra following this guide.
1. Log into Hydra web app at https://app.hydra.ai
2. Click on Scanners icon on the left navigation bar
3. Click on the Create New Scanner icon (+ icon on the bottom right of the screen)
4. You should see a list of available Scanner Templates. Click on the Search icon on the top right of the screen
Search for “box” to locate Scanner templates for Box
5. Identify the Metadata Extraction template, click SELECT
6. On the Scanner details page, type in Scanner Name and Description (e.g. Scanner name: Employee contract metadata scanner, Scanner description: Employee contract metadata scanner), and click on the NEXT button
7. On the Scanner Labels screen, type in label details, and click NEXT.
A label is a name that you use to recognize a data point that fits a specific pattern. As an example, if you are building a Scanner to extract names and dates from employment contracts, your labels are likely to include: employee name and contract date. If you are building a Scanner to categorize customers by their revenue potential, the labels are likely to include a ranking such as: Tier 01, Tier 02, Tier 03, etc.
8. On the Data Configuration screen, Provide the following input values, and click NEXT
Map to skill: Name of the custom Box Skill you have created for this Scanner ( how to create a Box Skill for Hydra). If this field is left empty, the Scanner will get invoked on any Box Skill.
Enterprise ID: Your Box workspaces' enterprise ID. If you are not planning on using a custom metadata template, type in: global
Metadata Template ID: Custom metadata template ID. If you are not planing on using a custom metadata template, type in: properties
Hydra will display the labels you created in the previous step. Assign them to the appropriate Box metadata fields. See the example below:
Contract Date Metadata Field Name: contract date
Employee name Metadata Field Name: employee name
9. On the Actions page, click NEXT
Actions provide the options to execute a task using the Scanner's output (i.e. predictions). Action configuration is optional. The available actions list is dynamically populated based on the Scanner Template and the connections available on your workspace.
10. On the Scanner Summary page, click on the Close icon on the upper-right corner to exit the Scanner configuration wizard.
Your new Scanner should appear on the Scanners page in draft mode. The next step is training and activating your Scanner using a sample document set.