> ## Documentation Index
> Fetch the complete documentation index at: https://yn-c9bb3266.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# AI-Powered Optimization

> Boost approval rates, reduce costs, and prevent fraud with Yuno's AI engine, Aida.

## Overview

Yuno's AI engine, **Aida**, uses machine learning to optimize payment routing, maximize approval rates, minimize processing costs, and detect fraud in real-time. Aida analyzes transaction patterns across Yuno's network to make intelligent decisions that improve your payment performance without requiring manual configuration.

## Core Capabilities

<div className="mdx-card-tiles">
  <CardGroup cols={2}>
    <Card title="Smart Routing" icon="route">
      Automatically route transactions to the provider most likely to approve them based on historical data, card type, issuer, and transaction characteristics.
    </Card>

    <Card title="Approval Rate Optimization" icon="chart-line">
      Increase authorization rates by selecting optimal providers, retry paths, and transaction parameters.
    </Card>

    <Card title="Cost Optimization" icon="coins">
      Minimize processing fees by routing to the most cost-effective provider while maintaining high approval rates.
    </Card>

    <Card title="Fraud Detection" icon="shield">
      Real-time fraud scoring that blocks suspicious transactions without adding friction for legitimate customers.
    </Card>
  </CardGroup>
</div>

## Smart Routing

Aida evaluates multiple factors in real-time to determine the best provider for each transaction:

| Factor                     | Description                          |
| -------------------------- | ------------------------------------ |
| **Card BIN**               | Issuer-specific approval patterns    |
| **Transaction amount**     | Provider performance by amount range |
| **Country/currency**       | Regional provider strengths          |
| **Payment method**         | Method-specific routing rules        |
| **Historical performance** | Provider approval rates over time    |
| **Time of day**            | Temporal patterns in approval rates  |

### How Smart Routing Works

```
Transaction arrives
    │
    ├── Aida evaluates all eligible providers
    │     ├── Provider A: 94% predicted approval, $0.30 fee
    │     ├── Provider B: 91% predicted approval, $0.25 fee
    │     └── Provider C: 87% predicted approval, $0.20 fee
    │
    └── Routes to Provider A (highest predicted approval)
```

<Info>
  Smart routing requires at least two payment providers configured for the same payment method and country. Aida continuously learns from each transaction to improve routing decisions.
</Info>

## Approval Rate Optimization

Beyond routing, Aida optimizes transaction parameters to increase approval likelihood:

* **Retry logic**: Automatically retries declined transactions on alternative providers when the decline is likely recoverable.
* **Parameter optimization**: Adjusts transaction fields (e.g., soft descriptors, MCC codes) to improve issuer approval.
* **Decline analysis**: Categorizes declines as soft or hard to determine retry eligibility.

## Cost Optimization

Aida balances approval rates against processing costs:

| Strategy           | Description                                                        |
| ------------------ | ------------------------------------------------------------------ |
| **Cost-first**     | Prioritize lowest fees while maintaining acceptable approval rates |
| **Approval-first** | Prioritize highest approval rates regardless of cost               |
| **Balanced**       | Optimize for best overall value (default)                          |

<Note>
  You can configure your optimization strategy in the Yuno Dashboard under **Settings > Smart Routing > Optimization Strategy**.
</Note>

## Fraud Detection

Aida's fraud detection engine scores every transaction in real-time:

| Risk Level | Score Range | Action                                   |
| ---------- | ----------- | ---------------------------------------- |
| Low        | 0-30        | Transaction proceeds normally            |
| Medium     | 31-70       | Transaction proceeds, flagged for review |
| High       | 71-100      | Transaction blocked or requires 3DS      |

### Fraud Signals

Aida evaluates hundreds of signals including:

* Device fingerprinting and behavioral analysis
* Velocity checks (transaction frequency per card/device)
* Geolocation consistency (IP vs. billing address)
* Known fraud patterns from Yuno's network

<Warning>
  Fraud detection rules can be customized in the Dashboard. Be careful when adjusting thresholds -- overly aggressive rules may block legitimate transactions, while lenient rules increase fraud exposure.
</Warning>

## Monitoring Performance

Track Aida's impact through the Dashboard analytics:

* **Approval rate trends**: Before and after enabling smart routing
* **Cost per transaction**: Average processing cost over time
* **Fraud metrics**: Blocked transactions, false positive rate
* **Provider performance**: Comparative analysis across providers

## Enabling Aida

Aida is available for merchants with multiple payment providers. To enable:

1. Navigate to **Settings > Smart Routing** in the Yuno Dashboard.
2. Enable **AI Optimization**.
3. Select your optimization strategy (cost-first, approval-first, or balanced).
4. Configure fraud detection thresholds if needed.

<Info>
  Aida requires a minimum transaction volume to build effective models. During the initial learning period (typically 2-4 weeks), routing may rely more heavily on default rules before Aida's predictions reach optimal accuracy.
</Info>