AWaws-samples
AI Assistant

POST /api/ai_assistant/converse

Changed on

POST operation for /api/ai_assistant/converse

post/api/ai_assistant/converse

Request

  • The document declares no server URL.
  • Auth
    socaAuth (a scheme the document does not define) and socaToken (a scheme the document does not define)

Headers

X-EDH-USERstring required

EDH username for authentication

X-EDH-TOKENstring required

EDH authentication token

Request body

promptstring required

The user message to send to the AI assistant

system_promptstring nullable

Optional system instruction to guide the AI response

max_tokensinteger nullable

Maximum number of tokens in the response

temperaturenumber float nullable

Temperature (0.0 = deterministic, 1.0 = creative)

top_pnumber float nullable

Top-p nucleus sampling parameter

knowledge_base_idstring nullable

Bedrock Knowledge Base ID for RAG-based responses

number_of_resultsinteger

Number of KB chunks to retrieve (only used with knowledge_base_id)

inference_profile_arnstring nullable

Custom Bedrock inference profile ARN. Defaults to cluster-configured model.

model_idstring nullable

Bedrock model ID to use for inference. Defaults to cluster-configured model.

Example request

{
  "prompt": "Why is my HPC job failing with mount errors?",
  "max_tokens": 4096
}

Response

AI assistant response generated successfully

successboolean

Example response

{
  "success": true
}

Changes