AI Methodology

Reviewora uses artificial intelligence to help customers turn their genuine experience into well-written review drafts. This page explains exactly how our AI works, what information it uses, and what it cannot do.

What AI Model We Use

Reviewora uses Google Gemini (specifically gemini-3.8-flash as the primary model with gemini-3.5-flash-lite as fallback) to generate review drafts. The AI runs entirely on the server side — no AI processing happens on the customer's device.

What Information the AI Receives

When a customer requests review drafts, we send the AI only:

What the AI Does

The AI generates three distinct review drafts in different tones (natural, friendly, and professional) based solely on the information provided by the customer. Each draft organizes the customer's own experience into clear, readable text.

What the AI Cannot Do

Customer Control

After the AI generates drafts, the customer has full control:

Data Privacy

Customer session data is sent to Google Gemini solely for generating review drafts. We do not use customer data to train AI models. See our privacy policy for full details.

Reliability

Reviewora implements a two-model cascade with automatic failover, retry logic, and circuit breaker protection. If the primary AI model is temporarily unavailable, the system automatically falls back to the secondary model. A total generation budget of 25 seconds ensures customers are never left waiting indefinitely.

Limitations

AI-generated drafts are suggestions, not guaranteed content. The quality of drafts depends on the information provided by the customer. We recommend customers always review and edit drafts before posting. AI models may occasionally produce imperfect phrasing.