About Feature

Intelligent Document Processing (IDP) reads the documents your suppliers and teams already send you — certificates of analysis, certificates of insurance, spec sheets, artwork, compliance files — and turns them into structured Specright records. Upload a PDF or image, choose what you want pulled from it, and IDP combines optical character recognition (OCR) with large language models to extract both individual values and full tables. Before anything is saved, you review the results side by side with the source document, then submit to create a new record or update one that already exists.

Because IDP reads what a document says rather than where things sit on a fixed template, it keeps working when a supplier changes their layout or a new vendor sends a format you haven't seen before.

Here's what IDP does for the different people who use it:

• End users create records straight from documents instead of keying in every field by hand.
• Reviewers and power users get a much faster verification step. Extracted text is highlighted on the document image, so checking a value against its source takes a glance.
• Administrators decide what gets extracted, where it's saved, and who reviews it, through configurable document types, field mappings, and permissions.
• Suppliers and partners can submit documents themselves through the supplier community, feeding your compliance workflows without an extra handoff.

If you've ever retyped a COA's test results into a Quality record line by line, or watched a spreadsheet of insurance expiration dates drift out of date because nobody had time to update it, this is built to remove that friction.

Feature Video Link(s)

IDP (Intelligent Document Processing) Self Service using LLM Release 27: 

https://youtu.be/wDoEs9BBwAo?si=YZaknSSBXegW2tsF

IDP Self Service Enhancements- Summer 2025 Release 30:  

https://youtu.be/j-OwygFmFIs?si=45t9FRoWA_S75BP6

Release 31 - Enhancements: 

https://www.youtube.com/watch?v=OjYP-4yUFWE&t=1s

Release 33 - Enhancements: 

https://www.youtube.com/watch?v=Ppu62vDh0pg

Application Area

IDP works from the Document Analysis tab, from a Document Extraction action on record pages, through email, and through bulk folder upload. Extracted data can be saved to any Specright object, standard or custom.

Ways to submit a document

• Document Analysis tab — upload a single file, or use Bulk Upload to send an entire desktop folder at once (without Salesforce's usual 10-file upload limit). Each file gets its own Document Analysis record and review task.
• Document Extraction action — start extraction from a record's detail page and write the results back to that same record.
• Email — send PDFs to a dedicated address set up for each document type. IDP processes them automatically and notifies the assigned reviewer.
• Supplier community — external supplier users submit documents directly.

Supported formats

PDF, PNG, JPEG, and TIFF. Documents need to be in English. Email submission accepts PDF files.

Two ways to process documents

• IDP Self Service (recommended) uses large language models and works with any document type you define. You choose which fields and table columns to extract, and you can pick between OpenAI and Amazon Nova models for each document type.
• In-house COA and COI models are machine learning models pre-trained on common certificate of analysis and certificate of insurance layouts. These come with a fixed set of default fields. You can add custom fields so they appear on the review screen, but IDP won't extract values for them automatically. For most COA processing, the Self Service route with the "Quality - COA" document type is now the better fit, since it doesn't depend on model training.

Core Benefits

• Turn hours of data entry into minutes — process documents one at a time, a folder at a time, or straight from an inbox, and let IDP do the typing.
• Handle whatever format arrives — IDP adapts to different layouts, so a new supplier template doesn't break your process.
• Review before anything is saved — extracted text is highlighted on the document image, and clicking a field or table cell boxes the matching value, so reviewers can confirm each one against its source.
• Create new records or update existing ones — choose Update Existing on the review screen to see each field's current value next to the extracted value, then check only the fields you want to overwrite.
• Capture tables, not just fields — define the columns you need, preview and edit the extracted table, and save its rows as related list records on a child object.
• Teach IDP your documents' vocabulary — add "also known as" labels so "Item Name" and "Product Description" land in the same field, and give the model document-specific instructions for tricky formats.
• Make document data reportable — once COA and COI data lives in Specright, you can track policy expirations, flag non-compliant certificates, and build dashboards on information that used to be locked in PDFs.
• Turn hours of data entry into minutes — process documents one at a time, a folder at a time, or straight from an inbox, and let IDP do the typing.
• Handle whatever format arrives — IDP adapts to different layouts, so a new supplier template doesn't break your process.
• Review before anything is saved — extracted text is highlighted on the document image, and clicking a field or table cell boxes the matching value, so reviewers can confirm each one against its source.
• Create new records or update existing ones — choose Update Existing on the review screen to see each field's current value next to the extracted value, then check only the fields you want to overwrite.
• Capture tables, not just fields — define the columns you need, preview and edit the extracted table, and save its rows as related list records on a child object.
• Teach IDP your documents' vocabulary — add "also known as" labels so "Item Name" and "Product Description" land in the same field, and give the model document-specific instructions for tricky formats.
• Make document data reportable — once COA and COI data lives in Specright, you can track policy expirations, flag non-compliant certificates, and build dashboards on information that used to be locked in PDFs.

Where is IDP available?

Release Updates

Level of Effort

IDP is configured by an admin, one document type at a time. End users don't need any setup of their own — once a document type is configured, they can start submitting documents.

For the admin, setup involves:

• Connecting your org to Specright's AWS-hosted extraction service (Named Credentials, plus API access enabled by the Specright product team)
• Assigning the right permission sets to internal users and, if needed, supplier community users
• Activating the IDP review task page and adding an Activities component to the Document Analysis record page
• Creating a document type and an IDP Self Service Setting for each kind of document you want to process, ideally with a fieldset so the right fields are pre-selected
• Optionally turning on email submission, bulk upload, the record-page Document Extraction action, field aliases, and custom extraction instructions

Once authentication is in place, expect roughly 2–3 hours to configure and test each Self Service document type, plus a short turnaround from the product team to enable API access. The in-house COI setup takes longer — plan for about half a day — because it needs two custom objects and a three-level parent, child, and grandchild mapping.

Short Setup Summary

1. IDP Review Task.
2. Configure AWS Named Credentials for ML model access.
3. Grant Specright API permissions and assign profiles.
4. Clone and configure Document Analysis record page.

5. Configuring IDP Self Service (Recommended)

   a. How to add a new Document Type

   b. Review Screen for IDP Self Service

   c. Setting Up Email Alerts for Review Tasks 

   d. Setting Up Email Submission Option for Self Service Document types

   e. IDP Bulk Upload setup (optional)

   f. IDP Self Service Settings

   g. Setting Up IDP “Document Extraction” on Record Detail Page

6. Configuring IDP for In-house COA Processing via Specright Settings > IDP Configuration.

7. Configure IDP for In-house COI Processing via Specright Settings > IDP Configuration

   a. COI components.

   b. Steps to configure IDP for COI 

   c. Recommended Setup to gain value from COI automation.

8. Using IDP.