ForjinnForjinn
User Experience

Onboarding Guide: Getting Started with Forjinn

This guide walks new users through their very first experience building, testing, and sharing smart workflows and agents in Forjinn. It aims to maximize success, minimize frustration, and accelerate the value journey from "hello world" to production.

Step 1: Sign Up or Accept an Invite

  • Choose "Sign Up" (admin: invite users via Workspaces)
  • Set a strong password or connect SSO (if available)
  • Complete profile; join or select your Workspace

Step 2: Platform Overview

The Forjinn dashboard features a tile-based layout for quick navigation to all major features:

  • Dashboard: Your home base with tiles for Chat Builder, Visual Canvas, API Gateway, Agents, Triggers, Datasets, Marketplace, and Settings
  • Workspace Selector: Pick your team/project from the top bar
  • Search & Help: Use the top bar to search docs or open the built-in help widget

Step 3: Choose Your Build Mode

Forjinn offers two primary ways to build AI workflows:

Chat Builder (Quick Start)

  • Streamlined, form-based interface for building conversational AI agents
  • Ideal for: Quick prototyping, chatbots, customer support agents, simple tool integrations
  • Select your LLM, configure tools, set system prompts, and deploy -- all without touching a canvas

Visual Canvas (Advanced)

  • Full drag-and-drop workflow builder with complete node customization
  • Ideal for: Complex multi-step workflows, RAG pipelines, conditional branching, advanced agent chains
  • Connect nodes, configure parameters, and build sophisticated automations

Tip: Start with Chat Builder for speed, then switch to Visual Canvas when you need more control.

Step 4: Create Your First Agent (Chat Builder)

  1. From the dashboard, click the Chat Builder tile
  2. Click New Agent
  3. Choose an AI framework:
    • Standard (LangChain): Great for single-agent tools and RAG
    • Google ADK: Leverage Google's Gemini models and ecosystem
    • CrewAI: Build role-based multi-agent teams
    • AutoGen: Create conversational multi-agent patterns
  4. Configure your agent:
    • Select your LLM provider and model
    • Add tools (search, API calls, Gmail, etc.)
    • Write a system prompt defining your agent's behavior
    • Configure memory for conversation context
  5. Test your agent in the built-in chat preview
  6. Save and deploy

Step 5: Build with Visual Canvas (Optional)

  1. Go to Chatflows > Click New Chatflow (or click the Visual Canvas tile)
  2. Drag & drop nodes from the library (e.g., Start, LLM, Retriever, Output)
  3. Connect with edges (arrows) to define data flow
  4. Click nodes to configure (choose model, enter API keys, set defaults)
  5. Save your flow with a clear name

Step 6: Test & Debug

  • Use the canvas "Run" or "Test" button to send a message or input
  • Watch output stream live; inspect variables mid-flow
  • Use logs/tracing panel for detailed debugging (see Testing & Debugging)
  • Chat Builder includes an integrated chat preview for real-time testing

Step 7: Add Data & Integrations

  • Import your knowledge (PDFs, CSV, JSON, docs) via Datasets or File Loader node
  • Add Credentials for external APIs (OpenAI, Google, Slack, etc.) via Settings
  • Try new tool/plugin nodes from the Marketplace

Step 8: Set Up API Gateway

  • Navigate to the API Gateway tile on your dashboard
  • Create endpoints to expose your agents and flows as REST APIs
  • Configure authentication, rate limiting, and request/response transformations
  • Share API documentation with your development team

Step 9: Automate with Triggers

  • Create Webhook Triggers to send notifications when flows complete
  • Set up Schedule Triggers to run batch jobs on a recurring basis
  • Configure Event Triggers to start workflows based on platform events

Step 10: Collaborate & Share

  • Add teammates to your Workspace
  • Share flows via export/import or project links (permissions-based)
  • Assign roles (admin, developer, viewer)

Step 11: Launch or Automate

  • Set up endpoints or webhooks for integration with external systems
  • Schedule runs or deployments via Marketplace templates
  • Monitor logs, usage, and quotas for your flows and agents

Step 12: Level Up

  • Explore advanced features (Multi-Agent with CrewAI, Google ADK, AutoGen)
  • Build complex RAG pipelines with custom datasets on Visual Canvas
  • Access Help Desk, Knowledge Base, and Community for ongoing guidance

Onboarding is just the start. Use this documentation portal for detailed walkthroughs, keep experimenting, and share your wins with the community!

Performance Optimization & Scaling

Previous Page

Forjinn Best Practices | Enterprise AI Development Guidelines

Essential best practices for building robust AI workflows in Forjinn. Security, scalability, and maintenance guidelines for enterprise AI development.

On this page

Step 1: Sign Up or Accept an InviteStep 2: Platform OverviewStep 3: Choose Your Build ModeChat Builder (Quick Start)Visual Canvas (Advanced)Step 4: Create Your First Agent (Chat Builder)Step 5: Build with Visual Canvas (Optional)Step 6: Test & DebugStep 7: Add Data & IntegrationsStep 8: Set Up API GatewayStep 9: Automate with TriggersStep 10: Collaborate & ShareStep 11: Launch or AutomateStep 12: Level Up