Skip to main content

Using the REST API

This guide provides practical examples for the most common tasks in uxopian-ai: managing conversations, sending requests, orchestrating goals, and accessing admin endpoints.

For details on the security model and authentication headers, see Security Model (BFF Pattern).

Prerequisites: Authentication Headers​

Uxopian-ai never handles authentication itself — it relies on a BFF Gateway to validate credentials and inject identity headers into every request. See Security Model (BFF Pattern) for the full architecture.

HeaderDescriptionRequired
X-User-TenantIdIsolates data per tenant (organization).Yes
X-User-IdUnique identifier for the user.Yes
X-User-RolesComma-separated list of roles (e.g., admin,user).No
X-User-TokenOriginal user token, forwarded to integrations (FlowerDocs, ARender).No

The Gateway authenticates the user (OAuth2, JWT, LDAP...), extracts identity from the session, and enriches the request with these X-User-* headers before forwarding it to uxopian-ai. The AI service trusts these headers implicitly — which is why uxopian-ai must never be exposed directly to the network.

Development Mode

In development mode (SPRING_PROFILES_ACTIVE=dev), the AI service accepts requests without headers and fills in defaults (User-development / Tenant-development). The examples below include explicit headers so they work in both dev and production environments. See Development Mode for details.

1. Creating a Conversation​

Conversations are the container for all history and context. They are scoped to the specific X-User-TenantId provided in the headers.

  • Endpoint: POST /api/v1/conversations

Example: Gateway forwarding a creation request​

cURL

curl -X POST "http://localhost:8080/api/v1/conversations" \
  -H "Content-Type: application/json" \
  -H "X-User-TenantId: enterprise-corp-a" \
  -H "X-User-Id: user-james-bond" \
  -H "X-User-Roles: user"

JavaScript (Node.js / Gateway Logic)

const createConversation = async (tenantId, userId) => {
  const response = await fetch(
    "http://uxopian-ai-service:8080/api/v1/conversations",
    {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
        "X-User-TenantId": tenantId,
        "X-User-Id": userId,
      },
    }
  );

  const data = await response.json();
  console.log("New Conversation ID:", data.id);
  return data;
};

2. Sending a Text Request (Non-Streaming)​

Send a prompt to the LLM within a conversation. The X-User-TenantId ensures the user only accesses conversations they belong to.

  • Endpoint: POST /api/v1/requests
  • Query Parameter: conversation (Required)

Example: Standard User Query​

cURL

curl -X POST "http://localhost:8080/api/v1/requests?conversation=123e4567-e89b-12d3-a456-426614174000" \
  -H "Content-Type: application/json" \
  -H "X-User-TenantId: enterprise-corp-a" \
  -H "X-User-Id: user-james-bond" \
  -d '{
    "inputs": [
      {
        "role": "user",
        "content": [
          {
            "type": "text",
            "value": "How do I reset my password?"
          }
        ]
      }
    ]
  }'

3. Triggering a Goal (Orchestration)​

Use the goal type to let uxopian-ai select the correct prompt based on the payload context.

  • Content Type: goal
  • Payload: Context data for the filter engine.

Example: Intelligent Goal Routing​

cURL

curl -X POST "http://localhost:8080/api/v1/requests?conversation=123e4567-e89b-12d3-a456-426614174000" \
  -H "Content-Type: application/json" \
  -H "X-User-TenantId: enterprise-corp-a" \
  -H "X-User-Id: user-james-bond" \
  -d '{
    "inputs": [
      {
        "role": "user",
        "content": [
          {
            "type": "goal",
            "value": "analyze_document",
            "payload": {
              "docType": "financial_report",
              "quarter": "Q3"
            }
          }
        ]
      }
    ]
  }'

4. Sending a Streaming Request​

For real-time responses (typing effect), use the streaming endpoint.

  • Endpoint: POST /api/v1/requests/stream
  • Response Content-Type: text/event-stream

Example: Streaming Response​

cURL

curl -X POST "http://localhost:8080/api/v1/requests/stream?conversation=123e4567-e89b-12d3-a456-426614174000" \
  -H "Content-Type: application/json" \
  -H "X-User-TenantId: enterprise-corp-a" \
  -H "X-User-Id: user-james-bond" \
  --no-buffer \
  -d '{
    "inputs": [
      {
        "role": "user",
        "content": [{ "type": "text", "value": "Explain quantum physics." }]
      }
    ]
  }'

5. Administrative Operations​

To access Admin endpoints, the Gateway must inject the admin role in the X-User-Roles header.

Example: Fetching Global Statistics​

cURL

curl -X GET "http://localhost:8080/api/v1/admin/stats/global" \
  -H "X-User-TenantId: enterprise-corp-a" \
  -H "X-User-Id: admin-user" \
  -H "X-User-Roles: admin"

Example: Listing LLM Provider Configurations​

cURL

curl -X GET "http://localhost:8080/api/v1/admin/llm/provider-conf" \
  -H "X-User-TenantId: enterprise-corp-a" \
  -H "X-User-Id: admin-user" \
  -H "X-User-Roles: admin"

Example: Creating an LLM Provider Configuration​

cURL

curl -X POST "http://localhost:8080/api/v1/admin/llm/provider-conf" \
  -H "Content-Type: application/json" \
  -H "X-User-TenantId: enterprise-corp-a" \
  -H "X-User-Id: admin-user" \
  -H "X-User-Roles: admin" \
  -d '{
    "provider": "openai",
    "defaultLlmModelConfName": "gpt5",
    "globalConf": {
      "apiSecret": "sk-your-api-key",
      "temperature": 0.7,
      "maxRetries": 3,
      "timeout": "60s"
    },
    "llModelConfs": [
      {
        "llmModelConfName": "gpt5",
        "modelName": "gpt-5.1",
        "multiModalSupported": true,
        "functionCallSupported": true
      },
      {
        "llmModelConfName": "gpt5-mini",
        "modelName": "gpt-5-mini",
        "temperature": 0.3,
        "multiModalSupported": true,
        "functionCallSupported": true
      }
    ]
  }'

For the complete list of LLM provider management endpoints, see REST API Reference — LLM Providers.

Last updated on