Quick Start

Set up your first AI workflow in Lasius step by step.

Get your first workflow running in Lasius. This guide provides an overview of the setup process with links to detailed guides for each step.

Prerequisites

  • A Lasius account (Sign up)
  • API key from at least one model provider (OpenAI, Anthropic, Google, or AWS Bedrock)

Setup Steps

StepRequiredDescription
0. Profile SetupYesConfigure your personal settings
1. Model ProviderYesSet up AI model access
2. Create WorkflowYesBuild your first workflow
3. MCP ServersOptionalAdd external tool capabilities
4. Knowledge BaseOptionalEnable document retrieval

Step 0: Set Up Your Profile

Before creating workflows, configure your personal settings:

  1. Click your profile icon in the sidebar
  2. Set your display name and preferences
  3. Configure notification settings for workflow alerts

Note:

Personalize AI Responses: Enable Global Prompt in Settings to set instructions, tone, and context that apply to all your AI agents and workflows. This ensures consistent, personalized responses across your entire workspace.

Learn more about Global Prompt →

Step 1: Configure a Model Provider

Workflows require an AI model to power agents. Set up at least one provider.

Supported providers: OpenAI, Anthropic, Google Gemini, Amazon Bedrock, Ollama

Quick steps:

  1. Go to Configurations in the sidebar
  2. Click Create New
  3. Select provider and enter API key
  4. Choose model and save

Note:

Configurations are reusable across all your workflows. Update once, apply everywhere.

Full Guide: Model Provider →

Step 2: Create Your First Workflow

Build your workflow using the visual Flow Editor.

Quick steps:

  1. Go to Workflows | Agents → Workflows
  2. Click New Flow
  3. Drag and drop: Trigger → Agent → Output
  4. Connect nodes and configure

Basic structure:

[Trigger] → [Agent] → [Output]

Full Guide: Workflow →

Step 3: Add MCP Servers (Optional)

Extend your agents with external tools via MCP (Model Context Protocol).

Capabilities: File access, database queries, API calls, browser automation

Quick steps:

  1. Navigate to MCP server configuration
  2. Add and configure your server
  3. Connect as a tool in your workflow

Full Guide: MCP Servers →

Step 4: Connect Knowledge Base (Optional)

Give your agents access to grounded information for RAG workflows.

Quick steps:

  1. Go to Knowledge Base in the sidebar
  2. Create and upload documents
  3. Add as a tool to your agent

Use cases: Q&A systems, research assistants, grounded responses

Full Guide: Knowledge Base →

Step 5: Run & Test

  1. Click Run in the workflow editor
  2. Monitor the execution trace in real-time
  3. Review outputs and iterate

Next Steps

Detailed Setup Guides

Advanced Topics