Parsium Documentation
Everything you need to install, configure, and use Parsium in your Salesforce org.
Configuration handled by Twilon
Most customers have Parsium configured by the Twilon team during onboarding, including Named Credential setup, permission assignments, model selection, and initial field mappings. If you prefer to set things up yourself, or want to understand how everything works under the hood, these guides walk you through every step so anyone can do it.
What is Parsium?
Parsium is an AI-powered document extraction tool built for Salesforce. It reads your documents (PDFs, images, spreadsheets, audio, video, and more), understands their content using AI, and automatically fills in the fields on your Salesforce records.
Upload Any File
Support for 15+ file formats including PDF, DOCX, XLSX, images, audio, and video.
AI Extracts the Data
Your chosen AI model reads the file and pulls out the exact information you need.
Fields Update Automatically
Extracted values are mapped to Salesforce fields and saved to the record instantly.
How It Works
Configure
Connect to OpenRouter, choose an AI model, and map your Salesforce fields.
Upload
Attach a document, image, or media file to any Salesforce record.
Extract
Parsium sends the file and your field instructions to the AI model.
Save
The AI response is parsed, validated, and saved to the record fields.
Prerequisites
- ✓A Salesforce org (Enterprise, Unlimited, or Developer Edition)
- ✓System Administrator or equivalent profile for initial setup
- ✓An OpenRouter account with an API key (free to create at openrouter.ai)
- ✓API credits on your OpenRouter account (pricing varies by AI model)
Documentation Sections
Setup & Installation
Start hereDeploy the package, create your Named Credential, configure API access, and assign permission sets.
Configuration
4 tabsConfigure the four settings tabs: credentials, privacy levels, AI model selection, and field mapping.
Field Mapping Guide
Key guideLearn how to create field mappings, write effective extraction instructions, and handle lookup fields.
Extraction & Files
15+ formatsStep-by-step extraction workflow, supported file types, and how each format is processed.
Reference
LookupTroubleshooting, system limits, quick reference tables, usage analytics, and FAQ.