SafeCloud_"Road to ${PRIZE_POOL}"

"Road to ${PRIZE_POOL}"

Comment

SafeCloud

Secure & Energy-Aware Cloud Operations Platform

SafeCloud is an AI-powered cloud operations platform that continuously monitors cloud environments for security vulnerabilities, resource inefficiencies, and carbon emissions. The platform combines rule-based analysis, customizable AI agents, and human-in-the-loop approvals to provide secure, sustainable, and explainable cloud optimization.

Project Overview

Modern cloud environments are becoming increasingly complex. Organizations must simultaneously balance:

  • Security and compliance
  • Cloud operational costs
  • Energy efficiency and carbon reduction
  • System availability and reliability

These objectives often conflict with one another. Increasing security may increase costs, while reducing costs may affect performance or operational resilience.

SafeCloud addresses this challenge by providing a unified platform that continuously scans cloud resources, evaluates findings through specialized AI agents, and recommends optimized actions while maintaining human oversight.

Key Features

Continuous Vulnerability Detection & Smart Remediation

SafeCloud continuously monitors cloud resources for security and operational risks, including:

  • Publicly accessible cloud storage
  • Unencrypted databases
  • Idle virtual machines
  • Unused storage resources
  • Resource misconfigurations

Detected issues are analyzed and prioritized before recommendations are generated.

Carbon Footprint Tracking & ESG Reporting

The platform utilizes cloud carbon emission data to:

  • Monitor carbon footprint trends
  • Track environmental impact
  • Estimate carbon savings from optimization actions
  • Generate ESG-ready sustainability reports

Users can export monthly ESG reports containing:

  • Carbon footprint trends
  • Emission reduction statistics
  • Optimization achievements
  • Sustainability metrics

Customizable AI Agents for Cloud Operations

SafeCloud employs a multi-agent architecture consisting of:

  • Security Agent
  • Energy Agent
  • Cost Optimization Agent
  • Risk Assessment Agent

Each agent analyzes findings from its own domain and contributes weighted recommendations to support decision-making.

Human-Approved Automated Actions & Audit Trail

To ensure accountability and governance, all recommended actions require human approval before execution.

The platform maintains:

  • Approval history
  • Action logs
  • Recommendation evidence
  • Audit trails

This human-in-the-loop approach provides transparency and reduces operational risks.

Architecture Overview

Cloud Resources
        ↓
Data Collection Layer
        ↓
Hybrid Scanning Engine
(Rule Engine + AI Analysis)
        ↓
Audit Validation Layer
        ↓
Master Agent
        ↓
Specialized AI Agents
        ↓
Recommendations
        ↓
Human Approval
        ↓
Action Execution & Audit Trail

Team Members

  • Ashley Chan Li Ling
  • Eugine Lee
  • Lau Zhe Hann
  • Loo Tan Yu Xian
  • Tan Yi Jie

Technologies Used

Frontend

  • TypeScript
  • React
  • Tailwind CSS

Backend

  • Python
  • FastAPI

Artificial Intelligence

  • Gemini

Database

  • PostgreSQL
  • Supabase

Deployment

  • Vercel
  • Render

Challenge & Approach

Why We Chose This Challenge

Cloud operations today require organizations to balance multiple competing objectives:

  • Security
  • Cost optimization
  • Energy efficiency
  • Sustainability

These priorities often contradict one another.

For example:

  • Increasing security controls may increase operational costs.
  • Reducing infrastructure costs may impact performance.
  • Improving system availability may increase energy consumption.
  • Lowering carbon emissions may require architectural changes.

Existing tools often focus on only one area, forcing teams to make decisions without understanding the broader impact.

Our team chose this challenge because we believe cloud operations should not optimize for a single objective. Instead, organizations need an intelligent system that evaluates security, sustainability, cost, and operational risk simultaneously.

SafeCloud was designed to bridge these competing priorities through AI-driven analysis, explainable recommendations, and human-centered decision making.

ESG Impact

SafeCloud contributes to Environmental, Social, and Governance (ESG) objectives through:

Environmental

  • Carbon footprint monitoring
  • Energy efficiency improvements
  • Resource optimization
  • Carbon reduction tracking

Social

  • Improved operational reliability
  • Reduced manual auditing effort
  • Better infrastructure visibility

Governance

  • Continuous compliance monitoring
  • Human approval workflows
  • Audit trails
  • Explainable AI recommendations

Future Improvements

  • Multi-cloud support (AWS, Azure, GCP)
  • Carbon-aware workload scheduling
  • Automated remediation workflows
  • Compliance framework support
  • Predictive cloud optimization
  • Construction-specific workload optimization
  • Advanced ESG reporting and analytics

Our Vision

SafeCloud aims to help organizations operate cloud environments that are:

  • Secure
  • Energy-efficient
  • Sustainable
  • Explainable
  • Governed

By combining AI-powered insights with human decision-making, SafeCloud enables organizations to optimize cloud operations responsibly while supporting both business objectives and ESG goals.

Built With

Share this project:

Updates