Skip to main content
This FAQ answers common questions about Veratrace based on implemented functionality. Questions about unimplemented features are marked with TODOs.

​
Authentication and Access

​
How do I authenticate to Veratrace?

Veratrace uses AWS Cognito for authentication via OIDC. Authentication configuration is fetched from your instance configuration. Source: Source: Feature Inventory

​
What roles are available in Veratrace?

Based on mock data, Veratrace has four roles:
  • Administrator - Full access (24 permissions)
  • Manager - Management and reporting (16 permissions)
  • Viewer - Read-only access (8 permissions)
  • Editor - Create and edit, no delete (12 permissions)
Source: Source: Feature Inventory

​
How do I create a new user?

Navigate to /users/create and fill in username, email, name, and temporary password. The user must change their password on first login. Source: Source: Feature Inventory

​
TWU Models

​
What is a TWU Model?

A Trusted Work Unit Model defines the structure for work operations. It specifies entities, actions, events, and outcomes that describe how work is performed. Source: Source: Feature Inventory

​
What are the TWU Model statuses?

TWU models have three statuses:
  • DRAFT - Unpublished, can be edited
  • PUBLISHED - Active and available for use
  • ARCHIVED - No longer active
Source: Source: Feature Inventory

​
How do I publish a TWU Model?

After creating and saving a draft, navigate to the model’s edit page and click the “Publish” button. This makes the draft version active. Source: Source: Feature Inventory

​
Can I have multiple versions of a TWU Model?

Yes. TWU models support versioning. You can create new versions, view all versions, and publish specific versions. Source: Source: Feature Inventory

​
Agents

​
What types of agents are supported?

Veratrace supports two agent types:
  • AI - Artificial intelligence agents with AI model configuration
  • HUMAN - Human agents with human-specific capabilities
Source: Source: Feature Inventory

​
What AI models are supported?

AI agents can be configured with models like “gpt-4”, “gpt-3.5-turbo”, and “claude-3”. The exact list depends on backend configuration. Source: Source: Feature Inventory

​
What capabilities can agents have?

AI agents have capabilities like text generation, code generation, image analysis, data analysis, translation, summarization, content creation, and automated responses. Human agents have capabilities like customer support, sales assistance, project management, strategic planning, relationship building, complex problem solving, creative thinking, leadership, training/mentoring, and quality assurance. Source: Source: Feature Inventory

​
Integrations

​
What integration types are supported?

Veratrace supports 40+ integration types across categories:
  • Contact Center/CX: AMAZON_CONNECT, GENESYS_CLOUD, ZENDESK, etc.
  • CRM/Sales: SALESFORCE, HUBSPOT, MICROSOFT_DYNAMICS, etc.
  • AI Agents: INTERCOM_FIN, MICROSOFT_COPILOT, etc.
  • Data/Documents: AWS_S3, SNOWFLAKE, etc.
  • And more…
Source: Source: Feature Inventory

​
How do I test an integration connection?

Navigate to /integrations/:integrationId/edit and click the “Test Connection” button. The system will verify connectivity and return test results. Source: Source: Feature Inventory

​
What are integration statuses?

Integrations can have the following statuses:
  • active - Operational
  • inactive - Disabled
  • connecting - Being set up
  • error - Encountered an error
Source: Source: Feature Inventory

​
Work Ledger

​
What is the Work Ledger?

The Work Ledger records all TWU executions and work operations. Each entry tracks execution status, policy compliance, actors (AI/HUMAN), cost, time, and related entities. Source: Source: Feature Inventory

​
What information is tracked in Work Ledger entries?

Each entry includes:
  • Title and description
  • Execution status (completed, in_progress, pending, failed)
  • Policy compliance status (compliant, non_compliant, review_required)
  • Actor types (AI, HUMAN, or both)
  • Active time (seconds)
  • Cost (dollars)
  • TWU model name
  • Related entity information
Source: Source: Feature Inventory

​
How do I filter Work Ledger entries?

Use the filter drawer on the /ledger page to filter by:
  • Search term (searches title, description, id, entityId, relatedEntity)
  • Status
  • Type
  • Policy status
  • Actors
  • Cost range
  • Time range
  • Model name
Source: Source: Feature Inventory

​
API and Development

​
What is the API base URL?

The development API base URL is http://localhost:8080. All endpoints are instance-scoped: /instances/:instanceId/{resource}. Source: Source: Feature Inventory

​
How do I authenticate API requests?

Source: Source: Feature Inventory

​
Are API responses paginated?

Yes. List endpoints return paginated responses with content, totalElements, totalPages, size, number, first, last, and numberOfElements fields. Source: Source: Feature Inventory

​
Instance Configuration

​
What is an Instance?

An Instance is a multi-tenant configuration container. Each instance has its own ID, AWS Cognito user pool configuration, and isolated data (users, TWU models, agents, integrations). Source: Source: Feature Inventory

​
How do I configure an Instance?

Source: Source: Feature Inventory

​
Evidence and Audit

​
How does evidence ingestion work?

Source: Source: Feature Inventory

​
What audit capabilities are available?

Work Ledger entries track policy compliance status for audit purposes. The system tracks compliance status (compliant, non_compliant, review_required) for each entry. Source: Source: Feature Inventory
  • Terminology - Key terms and definitions
  • Quickstart - Get started guide
  • Feature Inventory - See /docs/_generated/feature-inventory.md in the repository for complete feature reference