How Reinike AI Works

Complete technical documentation explaining the architecture, security model, and capabilities

1. Overview

What is Reinike AI?

Reinike AI is a desktop automation tool that transforms natural language instructions into executable Python code. Unlike cloud-based automation platforms, Reinike AI processes all files locally on your computer, ensuring complete privacy and data control.

The Hybrid Approach

Reinike AI combines the best of both worlds:

  • AI Intelligence: Natural language understanding via advanced language models
  • Local Privacy: All file operations execute on your computer
  • Instant Results: No upload queues, no waiting for cloud processing

2. Architecture Explained

Complete Technical Flow

Step 1: User types command in natural language

"Create 10 Excel files with cash flow tables"

Step 2: API Call - Only text instruction sent

Your instruction is securely transmitted to our AI service

Step 3: AI generates Python code

Step 4: Code validation and sanitization

Removes: pip commands, network calls, unsafe operations

Step 5: Local Execution - Runs on your PC

Python 3.12.1 bundled runtime

Step 6: Results appear in your folder

Execution time: 3-7 seconds average

3. Security Model

Files NEVER Leave Your Computer

Only your text instruction is sent to the AI service. The language model cannot see, read, or access your files in any way.

What Gets Sent to AI Service

  • ✅ Your text instruction
  • ✅ System prompt (safety rules)
  • ❌ NO file contents

4. Vision AI Integration

Image Understanding with AI Vision

Reinike AI includes a Vision AI feature that uses a multimodal language model to understand and extract information from images and scanned documents.

⚠️ Important: When using Vision AI, the image is sent to the Vision AI service. For maximum privacy with sensitive documents, use standard text-based mode.

5. Technical FAQ

Q: Can the AI access my files?

No. Only your text instruction is sent to the AI service.

Q: How secure is the generated code?

The system automatically blocks network requests and validates code before execution. Both Fast and Standard modes apply the same safety rules.

Q: What's the difference between Fast and Standard mode?

Fast Mode: Optimized AI model (2-5 seconds)

Standard Mode: Advanced AI model (5-10 seconds)

Both apply the same safety rules and validation.

Consumer Sovereignty OÜ
Kuusalu vald, Pudisoo küla, Männimäe/1,
74626, Estonia

Company

© 2025 Reinike AI. All Rights Reserved.