Quick Start

Get an AI agent running in four steps — create an agent, add a knowledge source, train it, then start chatting.

Step 1 — Create an Agent

Create your agent with a name and category. Save the id from the response — you'll need it in every subsequent step.

bash

curl -X POST 'https://api.7en.ai/api/v1/agents/' \
  -H 'Authorization: Api-Key YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "name": "Support Bot",
    "agent_category": "chatbot"
  }'

json

// Save the agent id from the response
{
  "data": {
    "id": 671,
    "name": "Support Bot",
    "status": "Idle"
  }
}

Step 2 — Add a Knowledge Source

Attach content to your agent so it has something to learn from. Supported types include url, text, and file.

Full reference →

bash

curl -X POST 'https://api.7en.ai/api/v1/knowledge-source/' \
  -H 'Authorization: Api-Key YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "agent": 671,
    "source_type": "url",
    "url": "https://your-site.com/help"
  }'

json

{
  "data": {
    "id": 42,
    "agent": 671,
    "source_type": "url",
    "status": "Pending"
  }
}

Step 3 — Train the Agent

Trigger training to index all attached knowledge sources. The agent's status immediately changes to "Training" and transitions to "Active" on success.

Full reference →

bash

curl -X POST 'https://api.7en.ai/api/v1/agents/671/train-agent/' \
  -H 'Authorization: Api-Key YOUR_API_KEY'

The endpoint streams back SSE events as training progresses:

text

training_started
{"message": "Agent training has started", "agent_id": "671", "status": "Training", ...}

training_training
{"train_data": {"phase": "extracting", "message": "Starting text extraction", ...}}

training_training
{"train_data": {"phase": "embedding_start", "message": "Starting embedding for 1 chunks", ...}}

training_active
{"agent_id": "671", "status": "Active", "train_data": {"phase": "embedding_completed", ...}}

Training time

Training typically completes within seconds to a few minutes depending on the size of the knowledge sources. Wait for status: "Active" before starting a chat session.

Step 4 — Chat via WebSocket

Connect to the WebSocket endpoint and send messages to start a real-time conversation with your agent.

Full reference →

bash

# WebSocket URL
wss://api.7en.ai/ws/api/chat/671/?api_key=YOUR_API_KEY

json

// Send a message
{ "message": "Hello, how can you help me?" }

// Receive a response
{
  "type": "bot_response",
  "content": "Hi! I'm your Support Bot. How can I help you today?",
  "session_id": "f5a48aab-9b43-4e88-b3a1-319475e385e0"
}

What's next

Authentication — learn how to manage and rotate API keys
Agent Operations — full agent configuration options
Knowledge Sources — all supported source types
Error Handling — understand error codes and responses

Ready to transform your business?

Start building intelligent AI agents to engage with customers.

← Authentication

Create Agent →

Home / API Reference / Quick Start

Quick Start

Get an AI agent running in four steps — create an agent, add a knowledge source, train it, then start chatting.

Step 1 — Create an Agent

Create your agent with a name and category. Save the id from the response — you'll need it in every subsequent step.

Full reference →

bash

curl -X POST 'https://api.7en.ai/api/v1/agents/' \
  -H 'Authorization: Api-Key YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "name": "Support Bot",
    "agent_category": "chatbot"
  }'

json

// Save the agent id from the response
{
  "data": {
    "id": 671,
    "name": "Support Bot",
    "status": "Idle"
  }
}

Step 2 — Add a Knowledge Source

Attach content to your agent so it has something to learn from. Supported types include url, text, and file.

Full reference →

bash

curl -X POST 'https://api.7en.ai/api/v1/knowledge-source/' \
  -H 'Authorization: Api-Key YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "agent": 671,
    "source_type": "url",
    "url": "https://your-site.com/help"
  }'

json

{
  "data": {
    "id": 42,
    "agent": 671,
    "source_type": "url",
    "status": "Pending"
  }
}

Step 3 — Train the Agent

Trigger training to index all attached knowledge sources. The agent's status immediately changes to "Training" and transitions to "Active" on success.

Full reference →

bash

curl -X POST 'https://api.7en.ai/api/v1/agents/671/train-agent/' \
  -H 'Authorization: Api-Key YOUR_API_KEY'

The endpoint streams back SSE events as training progresses:

text

training_started
{"message": "Agent training has started", "agent_id": "671", "status": "Training", ...}

training_training
{"train_data": {"phase": "extracting", "message": "Starting text extraction", ...}}

training_training
{"train_data": {"phase": "embedding_start", "message": "Starting embedding for 1 chunks", ...}}

training_active
{"agent_id": "671", "status": "Active", "train_data": {"phase": "embedding_completed", ...}}

Training time

Training typically completes within seconds to a few minutes depending on the size of the knowledge sources. Wait for status: "Active" before starting a chat session.

Step 4 — Chat via WebSocket

Connect to the WebSocket endpoint and send messages to start a real-time conversation with your agent.

Full reference →

bash

# WebSocket URL
wss://api.7en.ai/ws/api/chat/671/?api_key=YOUR_API_KEY

json

// Send a message
{ "message": "Hello, how can you help me?" }

// Receive a response
{
  "type": "bot_response",
  "content": "Hi! I'm your Support Bot. How can I help you today?",
  "session_id": "f5a48aab-9b43-4e88-b3a1-319475e385e0"
}

What's next

Authentication — learn how to manage and rotate API keys
Agent Operations — full agent configuration options
Knowledge Sources — all supported source types
Error Handling — understand error codes and responses

Ready to transform your business?

Start building intelligent AI agents to engage with customers.

← Authentication

Create Agent →