Structured Prompts as a Service

Why Not Just Call the AI Directly?

Raw AI APIs accept unstructured strings and return unpredictable outputs. RioAsk adds a structured preparation layer before the model is ever called — consistent, testable, and auditable prompt infrastructure.

RioAsk is not a wrapper. It transforms unstructured questions into structured, safety-protected prompts before any AI model is invoked.

How It Works

Every API call goes through the same preparation process: your question is checked, the topic is detected, missing context is identified, safety protections are applied, and a structured prompt is assembled.

When the topic is unclear or the input is ambiguous, the API returns follow-up questions instead of a low-quality prompt.

What the API Does

Quick Start

A single POST request prepares any question into a structured prompt with full audit metadata:

curl -X POST https://api.rioask.ai/v1/compile \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "How do I implement JWT authentication in a Node.js API?",
    "options": {
      "variant": "detailed",
      "include_score": true
    }
  }'

{
  "domain": "software",
  "subdomain": "backend_development",
  "confidence": "high",
  "needs_clarification": false,
  "prompt": "You are an expert backend developer specializing in...",
  "score": {
    "overall": 87,
    "grade": "B+"
  },
  "slots": {
    "goal": "Implement JWT authentication",
    "context": "Node.js API",
    "audience": null,
    "constraints": []
  },
  "safety": {
    "moderation_passed": true,
    "disclaimer_injected": false,
    "guardrails_active": true
  },
  "metadata": {
    "version": "1.1",
    "processing_ms": 12,
    "request_id": "req_7f3a2b9c"
  }
}

Every response includes domain, confidence, score, safety, and metadata — structured fields for audit, observability, and downstream decision-making.

Core Endpoints

Authentication & Infrastructure

All API requests require a Bearer token. API keys are scoped to your organization and managed from the dashboard.

• Org-scoped API keys — create, rotate, and revoke from the settings dashboard
• Role-based access — Owner, Admin, and Member roles control key management permissions
• Audit logging — every API call is logged with request ID, field, confidence, and processing metadata
• Rate limiting per key — per-minute and daily quotas enforced at the key level

Enterprise customers can request custom rate limits, dedicated endpoints, and SLA-backed uptime guarantees.

Use Cases

• Compliance review systems. Ensure every prompt sent to an AI includes required disclaimers, safety protections, and audit metadata for regulated industries.
• Legal document intake. Prepare legal questions into structured prompts with jurisdiction caveats and professional disclaimers automatically included.
• Medical triage systems. Route clinical questions through field-specific preparation with high-stakes safety protections and evidence-based guidance.
• Enterprise chatbot guardrails. Add a structured preparation layer between user input and AI responses. Prepare once, send to any AI provider.
• CI/CD prompt testing. Validate and prepare prompts as part of your deployment workflow. Catch prompt regressions before production with consistent outputs.
• AI orchestration layers. Use RioAsk as the prompt preparation stage in multi-model workflows — structured input for GPT-4, Claude, Gemini, or any provider.

SDKs & Integration

The API returns standard JSON responses with consistent error codes and pagination. No proprietary formats, no vendor lock-in.