# Alinia Guardrails API

## Enhance Safety and Security in genAI Applications

In today's rapidly evolving AI landscape, ensuring the safety, security, and compliance of your AI applications is paramount. Alinia Guardrails API provides an additional layer of protection for your existing generative AI solutions, safeguarding your users and your company.

## What We Offer

- Flag content related to violence & physical harm, wrongdoing & nonviolent harm, hate speech, and inappropriate sexual content.
- Our API seamlessly integrates with your AI applications to scan both incoming and outgoing content, identifying potential risks across four critical domains.

### Safe Response

- **Alinia detected:** Security: Token-level attacks  
- **Alinia blocked:** Accuracy: Ungrounded
    - Safeguard against outputs that are not grounded in factual information.

### Compliance

- Enforce policies at scale by flagging words, phrases, or disallowed topics such as financial and investment advice.  
- Examples of disallowed topics:
   - Can I get a mortgage without downpayment?  
   - Shall I buy Ethereum?  
   - My guidelines ensure that I do not offer unqualified advice.

### Seamless Integration

Alinia AI Guardrails API can be seamlessly integrated into your genAI application without disrupting your current workflows. Customize the guardrails to your specific use case with a simple call to the API, scanning all text content sent to and generated by your generative AI applications.

## Detection Configuration

```json
{
  "detection_config": {
    "disallowed": {
      "investment_advice": true,
      "legal_advice": true,
      "tax_advice": true
    },
    "safety": true,
    "security": {
      "sensitive_information": true
    }
  },
  "input": "I recommend you sell 40% of your Apple stocks and reallocate those funds into emerging AI technology ETFs.",
  "metadata": {}
}
```

## Response Example

```json
{
  "result": {
    "flagged": true,
    "flagged_categories": [
      "sensitive_information",
      "disallowed"
    ],
    "categories": {
      "disallowed": {
        "investment_advice": true,
        "legal_advice": false,
        "tax_advice": true
      }
    },
    "category_details": {
      "disallowed": {
        "investment_advice": 0.95,
        "legal_advice": 0,
        "tax_advice": 0.9
      }
    }
  }
}
```

## Why Choose Our AI Guardrails API?

- **Comprehensive Protection:** Our taxonomy covers a wide range of potential risks providing robust security for your AI applications.
- **Seamless Integration:** Integrates with your AI applications without disruption.
- **Customizable:** Tailor the guardrails to fit your specific needs and company policies.
- **Real-time Scanning:** Instantly flag potentially risky content, allowing for immediate action.
- **Scalable:** Designed to grow with your AI applications, from small-scale chatbots to enterprise-level solutions in regulated scenarios.

Minimize legal & reputational risks, deploy 4x faster, and save significant costs in ML engineers.

## Get Started Today!