How to Optimize Processed Page Count and Prevent Duplicates in DigitAI
Learn how DigitAI calculates processed page consumption for billing, how the duplicate file detection mechanism works, and best practices to audit and optimize your monthly usage.
Step by step
- 1
How DigitAI Calculates Processed Pages
In DigitAI, two distinct metrics define document length: 1. **Total Pages (`total_pages` / `pages_number`):** Represents the total physical page count of the uploaded file (e.g., a 3-page PDF). 2. **Processed Pages (`processed_pages`):** Represents the billable page count computed by the platform. #### Calculation and Billing Logic: - When the extraction engine processes a page using AI models, it assigns an internal flag `ai_extracted: true`. - The `processed_pages` metric counts only pages where `ai_extracted` is true. - If no pages receive this flag (e.g., non-AI extraction rules or disabled extractor), the system defaults to the total physical page count (`pages_number`) as a fallback. #### Multi-Page Documents with Identical Copies (Original, Duplicate, Triplicate): - **Without internal page deduplication:** If a document has 3 pages and the account extractor does not have page deduplication enabled, all 3 pages are processed and billed. - **With internal page deduplication:** When enabled, the extractor identifies identical subsequent pages, processes only the first page, and marks copies as `ai_extracted: false`. In this scenario, only 1 page is counted toward billing.
- 2
Automated Duplicate Document Detection (`check_duplicated`)
DigitAI provides a built-in mechanism to prevent duplicate processing: - **Default behavior:** Duplicate detection is disabled by default. Uploading the same file twice results in duplicate processing and double billing. - **Activation:** The setting (`check_duplicated: true`) is configured per customer. Contact support to enable it for your organization. - **How it works:** When enabled, the system computes the SHA-256 checksum of incoming files. If a match is found within your organization's records: 1. Sets document lifecycle status to `DUPLICATED` (`extractor = 'duplicated'`). 2. Links the new record to the existing original document and copies its metadata/tags. 3. Bypasses AI extraction pipelines. 4. **Does not add billable pages to your monthly quota.**
- 3
What Happens to Deleted Documents?
- Deleting a document from the platform **does not reduce or refund** processed page credits. - Once a document is processed by the AI engine, page consumption is permanently recorded in the billing period metrics.
- 4
Monitoring, Auditing, and Operational Best Practices
#### Configuring Document Grid Columns for Auditing: Both admins and standard users can customize the table view: 1. Navigate to the **Documents** section. 2. Open the column customization menu. 3. Enable the following columns: - **Total Pages** - **Processed Pages** - **Lifecycle:** Quickly identify documents marked as `DUPLICATED` or with processing issues. #### Usage Metrics Report: Administrators can audit consumption under the **Usage Metrics / Reports** section to review processed page trends throughout the billing cycle. #### Best Practices to Optimize Consumption: - Check existing documents using platform search filters prior to uploading new batches. - Request page deduplication configuration from support if your organization frequently handles multi-copy invoices. - Enable `check_duplicated` if files arrive via multiple channels (email, manual upload, API) to prevent accidental duplicate processing.
Last updated: 11/13/2026
Still have questions?
Contact support