Skip to main content
This guide provides a minimal overview of how to start using Avigrah to run AI workflows within a governed, production-ready environment.

​
1. Create an Organization

An Organization is the top-level isolation unit in Avigrah.
  • Represents a single company or client
  • Owns all projects, data, pipelines, credentials, and usage
  • Ensures complete data, execution, and billing isolation
All resources in Avigrah exist within an organization.

​
2. Invite Collaborators

Invite users to collaborate within your organization.
  • Roles define what actions each user can perform
  • Access is enforced across projects, pipelines, and AI execution
  • Users can only access resources within their assigned scope
Permissions are enforced at runtime, not just in the UI.

​
3. Create a Project

A Project defines a specific AI use case and its execution boundary.
  • Select an AI Template (e.g., Customer Intelligence, Ads Conversion, Investor Intelligence)
  • Defines scope for:
    • Data pipelines
    • Vector stores
    • AI agents
    • Teams
Each project ensures AI operates only on approved data.
One project = one AI workload.

​
4. Create a Team

Teams define who can access and use the AI.
  • Add users to teams
  • Assign teams to projects
  • Team members can only access outputs within assigned projects
Teams represent AI consumers, not infrastructure owners.

​
5. Create a Data Pipeline

Pipelines prepare data for AI execution.
  • Supports batch and real-time data sources
  • Executes in ordered phases:
    Ingest → Transform → Store → Vectorize → Ready for AI
  • Outputs are stored in project-scoped storage and vector databases
AI agents consume processed pipeline outputs, not raw data.

​
6. Create an Integration

Integrations connect external systems to Avigrah.
  • Configure third-party or internal system credentials
  • Credentials are stored at the organization level
  • Assign integrations to specific projects
Pipelines and agents can now access live operational data securely.

​
You’re Ready To Go

Your environment is now fully configured.
  • Data pipelines are active
  • AI agents can execute securely
  • Access is controlled via roles and teams
  • Usage is tracked per organization and project
You can now start generating insights, automation panel, and AI-driven workflows.