"Automating Compliance & Security for Modern DevSecOps Pipelines"
CodeGuard AI is an enterprise-grade security tool designed to detect GDPR violations, Cloud Security risks, and PII leaks directly in the source code. Powered by Google's Gemini 2.0 Flash LLM, it acts as a Senior Security Auditor, providing context-aware analysis and auto-remediation code fixes.
🚀 Key Features
🧠 Context-Aware AI Analysis
Unlike traditional regex-based scanners (SAST), CodeGuard understands the intent of the code.
False Positive Reduction: Distinguishes between a variable named password and an actual hardcoded credential.
Multi-Language Support: Analyzes Java, Python, Node.js, Terraform, and more.
🇪🇺 GDPR & Privacy Specialist
Specifically trained to detect violations of EU General Data Protection Regulation:
Data Minimization: Detects unnecessary logging of PII (User IDs, Emails, NIFs etc).
Data Sovereignty: Flags cloud configurations (AWS/Azure) that export data outside the EU.
✨ Magic Fix (Auto-Remediation)
Don't just find the bug—fix it.
One-Click Fix: Generates secure, refactored code snippets ready to copy-paste.
Educational Feedback: Explains why the code was insecure.
Executive Dashboard
Security Score: Real-time calculation of risk metrics based on AI findings.
Responsive Design: Enterprise-grade UI fully functional on Mobile and Desktop.
Multi-Language Report: Generates audits in English, Portuguese, Spanish, French, Italian, and German.
System Architecture
The project follows a distributed Microservices Architecture to ensure scalability and separation of concerns.
graph LR User[User / CI-CD] -- HTTP --> Front[Angular Frontend] Front -- REST API --> Back[Java Spring Boot] Back -- Internal API --> Engine[Python AI Engine] Engine -- Secure Request --> Gemini[Google Gemini 2.0]
Orchestration: Manages requests between Client and AI.
Smart Caching: Implements in-memory hashing to cache analysis results (reducing AI costs and latency to 0ms for repeated scans).
Resilience: WebFlux WebClient with timeouts and error handling.
AI Engine (Python Flask):
Prompt Engineering: Specialized personas for Security Auditing using Gemini 2.0.
Resilience: Handles Rate Limits (429) and API failures gracefully.
Universal Scanner: Language-agnostic analysis logic.
Techs
Frontend Angular 17, TypeScript Tailwind CSS (CDN), ngx-markdown, Responsive Layout Backend Java 17, Spring Boot 3 WebFlux, Maven, ConcurrentHashMap Cache AI Engine Python 3.10 Flask, Google GenAI SDK, Dotenv
Clone the repository:
git clone https://github.com/silvatechf/codeguard-ai.git cd codeguard-ai
"The dashboard provides a clear, actionable report for non-technical auditors, bridging the gap between Legal (DPO) and Engineering."
🛡️ Security & Privacy
This tool follows Privacy by Design principles:
Stateless Analysis: Source code is processed in-memory and discarded immediately after analysis.
No PII Storage: We do not store user data, only the analysis metadata.
Secure Transport: All internal communication is designed to run within a private Docker network.
👤 Author
Fernando Silva Software Engineer | Java & Angular Specialist | AI Enthusiast
LINKEDIN https://www.linkedin.com/in/fernando-silva-83b155a4/
© 2025 CodeGuard AI. All rights reserved.