Skip to main content

Overview

This quickstart guide will help you:
  1. Create an account and set up your organization
  2. Build your first agent using the web interface
  3. Test your agent in the chat interface
  4. (Optional) Integrate via API using the OpenAI-compatible endpoint
Prerequisites: A Google account for authentication. No coding required for steps 1-3!

Step 1: Create Your Account

1

Sign Up

Go to junis.ai and click “Sign in with Google”
2

Complete Onboarding

You’ll be redirected to the onboarding page. Fill in:
  • Organization Name: Your company or personal workspace name
  • Display Name: How you want to be addressed
3

Access Dashboard

You’ll land on the main dashboard. This is your command center!

Step 2: Create Your First Agent

1

Navigate to Orchestration

Click “Team Management” dropdown in the Navigation Bar, then select “Agent Orchestration”
2

Create New Agent

Click the “New Agent” buttonFill in the form:
Name: MyFirstAgent
Type: LLM Agent
Model: claude-haiku
Description: A helpful AI assistant
3

Write Instructions

In the “Instructions” field, write:
You are a helpful AI assistant. When users ask questions:
1. Provide clear, concise answers
2. Use friendly, professional tone
3. Ask clarifying questions when needed

Always be helpful and informative.
Good instructions are key to agent performance! Be specific about behavior and tone.
4

Save Agent

Click “Create” to save your agent
Agent created successfully!

Step 3: Test Your Agent

1

Open Chat Interface

Click “Chat” in the sidebar
2

Send a Message

Type a message and hit Enter:
Hello! What can you help me with?
Watch as your agent responds in real-time! 🚀
3

Explore Features

Try these features:
  • File Upload: Click 📎 to attach documents
  • Voice Input: Click 🎤 to speak your message
  • Session History: See past conversations in the sidebar

Step 4: Integrate via API (Optional)

Now let’s call your agent programmatically using our OpenAI-compatible API.
1

Generate API Key

  1. Navigate to “Dashboard”“API Keys”
  2. Click “Create New API Key”
  3. Enter:
    • Name: My First API Key
    • Description: Testing quickstart
    • Scopes: (leave default)
  4. IMPORTANT: Copy the key immediately - it won’t be shown again!
2

Test with cURL

Open your terminal and run:
curl -X POST https://api.junis.ai/api/external/v1/chat/completions \
  -H "X-API-Key: YOUR_API_KEY_HERE" \
  -H "Content-Type: application/json" \
  -d '{
    "messages": [
      {"role": "user", "content": "Hello from API!"}
    ],
    "stream": false
  }'
You should receive a response from your agent!
3

Enable Streaming

For real-time token-by-token responses:
curl -N -X POST https://api.junis.ai/api/external/v1/chat/completions \
  -H "X-API-Key: YOUR_API_KEY_HERE" \
  -H "Content-Type: application/json" \
  -d '{
    "messages": [
      {"role": "user", "content": "Tell me a story"}
    ],
    "stream": true
  }'

Next Steps


Common Issues

Solutions:
  1. Check if agent is active (toggle in admin panel)
  2. Verify LLM model is available (check status page)
  3. Refresh the page and try again
  4. Check browser console for errors (F12)
Solutions:
  1. Ensure you’re using the correct header: X-API-Key
  2. Check if key is still active in admin panel
  3. Verify you’re calling the correct endpoint: /api/external/v1/chat/completions
  4. Check for typos in the API key
Possible causes:
  • Model: Claude Opus is slower than Haiku. Switch to Haiku for faster responses
  • RAG: Large knowledge bases increase latency. Optimize document size
  • MCP: External service calls add time. Use caching when possible
  • Network: Check your internet connection

Need Help?