Inspiration
The project was inspired by the lack of readily available tools to perform quick and intelligent auditing of the environment with clear and concise reports that show actions to be taken to become compliant. While sophisticated enterprise solutions exist they are often complex to set up and expensive to run. This solution proves that results can be achieved quickly and cheaply for small to medium sized organizations that lack the required expertise to set up complex auditing solutions.
What it does
The solutions runs and evaluates environment scans to identify non-compliant resources while providing quick steps to resolve the issues. The tool is configurable for various compliance frameworks and provides a simple UI for observability, configuration and historical overview.
How we built it
The solution was built by utilizing a serverless and modular project that leverages modern methodologies to achieve low cost and impressive scalability while consistently maintaining high quality of results. It employs cloud native architectures, automated workflows, and efficient resource management to minimize operational overhead, enable seamless scaling under varying loads, and ensure reliability, security, and performance that meet enterprise grade standards across diverse use cases.
Core Capabilities
- Autonomous Compliance Auditing: AI-powered analysis of AWS resources against compliance frameworks
- Multi-Framework Support: GDPR, SOC 2, PCI-DSS compliance checking
- Intelligent Findings: Prioritized risk assessment with actionable remediation steps
- Automated Report Generation: Professional PDF reports with executive summaries and detailed findings
- Real-time Dashboard: Interactive React dashboard with compliance metrics and trends
- Secure Authentication: AWS Cognito integration for user management ## Technical Features
- Serverless Architecture: Zero idle cost with AWS Lambda, API Gateway, DynamoDB, and S3
- Amazon Bedrock Integration: Advanced AI reasoning
- Bedrock AgentCore Runtime: Enterprise-grade agent orchestration with memory and tools
- RESTful API: Comprehensive API for audit management and findings retrieval
- CloudWatch Monitoring: Built-in logging, metrics, and alerting
- Event-Driven Scanning: Automated scheduled compliance scans
Challenges we ran into
The main challenge was the automation of resource deployments that are related to the Bedrock AgentCore stack, especially since there is limited documentation available as it is a very new framework still in its early adoption phase. This scarcity of comprehensive guides and examples required extensive experimentation, reverse engineering of sparse API responses, and iterative testing to uncover undocumented behaviors, integration nuances, and potential pitfalls in configuration workflows. Despite these hurdles, it fostered innovative problem solving, deeper insights into the framework's underlying mechanics, and the creation of custom scripts that not only resolved the immediate deployment issues but also established a reliable blueprint for future automations, paving the way for more efficient and resilient infrastructure management in emerging AI driven environments.
Accomplishments that we're proud of
I am very proud of the resulting solution because it turned out even better than I hoped, delivering robust performance and elegant simplicity in every aspect. I am looking forward to continuing the work on it in the future, building on this strong foundation with new features, thoughtful refinements, and further innovations that will enhance its capabilities and reach.
What we learned
I have gained a lot of valuable experience and in depth knowledge related to the new SDKs and frameworks that are utilized in this project, mastering their core functionalities, exploring advanced integration techniques, and understanding best practices for optimal performance and scalability. I have become proficient in leveraging their unique capabilities to solve complex challenges, debugging intricate issues efficiently. This hands on exposure has significantly strengthened my technical expertise, deepened my appreciation for cutting edge tools, and prepared me to contribute even more effectively to future iterations, related initiatives, and broader engineering efforts with confidence, precision, and a clear vision for continued growth and innovation.
What's next for SecureAuditAI
I am looking forward to optimizing and expanding the features of this solution in thoughtful and strategic ways to bring the best possible results to small and medium size organizations. I plan to refine performance, enhance usability, and introduce powerful new capabilities that address real world needs, empower teams, and drive meaningful impact for businesses of all scales within this important segment.
Built With
- amazon-web-services
- bedrock
- cloudformation
- css
- github
- javascript
- python
- typescript
Log in or sign up for Devpost to join the conversation.