Skip to main content
Parsium Documentation

Configuration

The Parsium Settings page has two main tabs. Open it by clicking the Parsium Settings tab in your Salesforce navigation. What you see depends on whether you are an Owner or an Admin.

Owners vs Admins

Parsium separates two roles with different levels of access to settings.

OwnerThe person who connected OpenRouter. Stored in Protected Custom Metadata (invisible to Salesforce admins). Owners control the AI connection, model selection, privacy settings, voice configuration, and access management. All sensitive actions require PIN entry.
AdminA user with the Parsium_Admin permission set. Admins can manage field mappings, view OpenRouter service status, and configure analytics retention. Admins cannot change the AI model, privacy settings, or disconnect OpenRouter unless they are also an owner.

Field Mappings Tab

Tab 1Admin + Owner

Define which Salesforce fields should be populated by the AI and what instructions to give it. See the Field Mapping Guide for a full walkthrough.

  • Object SelectorA searchable dropdown listing every Salesforce object you have permission to update. Select the object you want to create mappings for.
  • Field TableAfter selecting an object, a table appears showing your current field mappings. Each row has a Field, Type badge, Instruction, and Delete button.
  • Add FieldClick "Add Field" to add a new mapping row. Select a field from the dropdown, then write an extraction instruction. Already-mapped fields are excluded from the dropdown.
  • Save AllSaves all field mappings for the selected object in a single batch. You must click Save after making changes.
  • Analytics RetentionConfigure how many days of Request Log data to retain. Available to admins.

OpenRouter Tab

Tab 2

The OpenRouter tab has two sub-tabs. What you see depends on your role.

Configuration Sub-tab

Owner Only

Only visible to owners. Every change in this sub-tab requires PIN entry. Contains all sensitive settings organized into sections:

  • ConnectionConnection status, API key label, usage statistics, and a disconnect button.
  • Model & ParametersThree model selectors: Extraction, Insight, and AI Mapping (Professional only). Plus reasoning effort toggle. See AI Models & Parameters below.
  • PrivacyPrivacy level toggle (Standard/Enhanced/Maximum), request logging checkbox, and show model name option. See Privacy Levels below.
  • InsightInsight data source: choose between Mapped fields only (uses your configured field mappings) or All populated fields (sends all non-blank field values to the AI).
  • CreditsView your OpenRouter credit balance and usage activity. Requires a Management API key (see below). Links to add credits on OpenRouter.
  • Management KeyOptional OpenRouter Management API key for viewing credits and usage. Guided paste setup. Stored in Protected Custom Metadata alongside your main API key.
  • Access ManagementView current owners, add co-owners, remove owners (immediate, with PIN verification), change your PIN, and view the audit log.
  • ThresholdsAuto-accept threshold and monthly cost threshold settings.

Status Sub-tab

Admin + Owner

Visible to both admins and owners. Shows the health of the OpenRouter connection and service.

  • Service StatusLive indicator (green/yellow/red) pulled from status.openrouter.ai with a description of current conditions.
  • Connection Health"Test Connection" button that verifies Parsium can reach OpenRouter with the stored API key.
  • Status Page LinkDirect link to the full OpenRouter status page for detailed incident information.

Parsium is an offline service. OpenRouter processes AI requests on behalf of Parsium. If OpenRouter shows degraded performance, extractions may be slower or temporarily unavailable.

PIN System

Every sensitive operation requires entering a 4-6 digit numeric PIN. The PIN is set during initial setup and stored as a SHA-256 hash with a random salt in Protected Custom Metadata.

  • When RequiredChanging the AI model, changing privacy settings, disconnecting OpenRouter, adding/removing owners, changing the PIN, and saving the management key.
  • Changing PINEnter your current PIN and a new PIN. Optionally generate a new recovery code at the same time.
  • Forgot PINUse your recovery code to reset the PIN. If you have lost both, contact Twilon support for a remote reset.

Recovery Options

Parsium provides two levels of recovery if you forget your PIN.

Recovery Code

A 16-character code (format: PRSM-XXXX-XXXX-XXXX) shown once during initial PIN setup. Store it securely. It allows you to reset your PIN without contacting support.

Remote Reset

If you lose both your PIN and recovery code, contact support@twilon.com. Twilon support can trigger a remote PIN reset through the License Management App. You will then set a new PIN and receive a new recovery code.

Audit Log

Every configuration change is permanently recorded in the Audit Log. Records are immutable: they cannot be edited, and they cannot be deleted within 90 days. This is enforced by a managed trigger that subscribers cannot disable.

  • What is LoggedModel changes, privacy changes, reasoning effort changes, Insight settings, owner additions/removals, PIN changes, recovery resets, connect/disconnect events, and management key changes.
  • Record FieldsChange type, old value, new value, who made the change, and when.
  • AccessReadable by admins and owners. Nobody can create, edit, or delete audit records manually. Only Parsium's internal service can create them.
  • Retention90-day minimum. Records older than 90 days can be deleted if needed, but the default behavior preserves them indefinitely.

Privacy Levels

Parsium offers three privacy levels that control how AI providers handle your data. The default is Maximum, the most restrictive option. Changing the privacy level requires owner access and PIN entry.

Standard

Level 0Most models available

No restrictions on data handling. AI providers may use your data according to their standard terms, which could include training future models. This gives you access to the widest selection of AI models at the lowest cost.

When to use: Use this for non-sensitive data where cost and model selection are more important than data privacy. Good for testing and development.

Enhanced

Level 1Good model selection

Parsium sends a data_collection: "deny" flag with every request, telling the AI provider not to use your data for training. Most major providers honor this flag. You still get a good selection of models.

When to use: Use this for business data that is not highly regulated but should not be used to train AI models. A good middle ground for most organizations.

Maximum

Level 2ZDR endpoints only

Everything from Enhanced, plus Parsium only shows models available through Zero Data Retention (ZDR) endpoints. ZDR means the provider guarantees your data is never stored, logged, or retained after the response is generated.

When to use: Use this for regulated industries (healthcare, finance, legal), sensitive documents, or any time data retention by a third party is unacceptable. This is the default setting.

What happens to your data when it's sent to an AI model?

Peer-reviewed research from Google DeepMind, Meta, and leading AI security researchers shows that your data is statistically impossible to memorize or retrieve from a trained model. We wrote a detailed breakdown with 14 cited sources.

Read: Is Your Data Safe with AI?

AI Models & Parameters

Parsium connects to AI models through OpenRouter, which gives you access to 200+ models from leading AI providers. Model selection is owner-only and requires PIN entry.

Choosing a Model

FactorWhat It Means
File supportNot every model can process every file type. Use the capability filters (Files, Images, Audio) to find models that support your file types.
Context lengthHow much text the model can handle in a single request. Large documents need models with 100k+ tokens.
PricingModels charge per token (~4 characters = 1 token). Smaller models are much cheaper than larger ones.
QualityLarger models produce more accurate extractions for complex documents. For simple invoices, a small model works fine.
SpeedSmaller models respond faster. If processing time matters, choose a smaller model.

Recommended starting point: The default model supports images, has a large context window, and offers a good balance of quality, speed, and cost. Switch to a smaller model for faster, cheaper extractions on simple documents.

Reasoning Effort

Some AI models support "reasoning", which means they think step-by-step before giving an answer. This improves accuracy for complex tasks but takes longer and costs more.

LevelValueDescription
Off0No reasoning. Standard response behavior. Fastest and cheapest.
Low1Minimal reasoning. Fastest response, lowest cost.
Medium2Moderate reasoning. Good balance of accuracy and speed.
High3Maximum reasoning. Slowest but most thorough for complex documents.

Models That Support Reasoning

New reasoning models are released frequently. Parsium automatically detects which models support reasoning and shows the effort toggle only when applicable. You can browse the current list of reasoning-capable models on OpenRouter:

View reasoning models on OpenRouter

Email Notifications

Parsium sends automatic email notifications to all owners when sensitive events occur. These cannot be disabled.

EventRecipients
AI model changedAll owners
Privacy level changedAll owners
Owner addedAll owners + new owner
Owner removedAll owners + removed user
PIN changedAll owners
OpenRouter disconnectedAll owners
Recovery code resetAll owners