CS 79D Final Project
Production-style AWS application stack mapped directly to the final assessment requirements: compute, storage, scaling, security, monitoring, and AI integration.
Final Answer
6 / 6 requirements covered
Evidence is gathered from the deployed app, CloudFormation, production API checks, and AWS service integrations.
Frontend
Next.js
Backend
NestJS API
AI Mode
Bedrock
The project is a cohesive frontend, backend, compute, storage, security, monitoring, and AI stack rather than isolated AWS demos.
Application Layer
Compute and Processing
Persistent Storage
Network and Delivery
Security Controls
Monitoring and Cost
AI and Automation
Each requirement is answered with the implementation, AWS services used, and current evidence status.
Complete
Answer
EC2 runs the Next.js frontend and NestJS backend, while Lambda handles event logging.
Verification
Production health and Lambda invocation checks returned successful responses.
Evidence
Complete
Answer
The stack uses two persistent AWS storage services: S3 for uploads and DynamoDB for activity records.
Verification
A test file persisted in S3 and corresponding file-upload records appeared in DynamoDB.
Evidence
Deployed
Answer
CloudFormation deploys an Application Load Balancer, frontend/backend target groups, and an Auto Scaling Group.
Verification
The production architecture uses ALB traffic distribution and ASG capacity management; final screenshots document healthy targets and desired capacity.
Evidence
Strong
Answer
Security is covered across IAM, security groups, HTTPS, headers, auth, upload validation, throttling, and bot checks.
Verification
Production checks confirmed security headers, JWT protection, Turnstile enforcement, and unsafe upload rejection.
Evidence
Implemented
Answer
CloudWatch logs, health checks, CPU alarms, dashboard widgets, and cleanup/cost notes support operations and cost control.
Verification
The health endpoint and Lambda log group are live; dashboard and alarm screenshots complete the operational evidence.
Evidence
Complete
Answer
Amazon Bedrock powers a live AI Security Advisor that generates rubric-focused architecture recommendations.
Verification
Advisor refreshes produce dynamic recommendations from Amazon Nova Micro, a low-cost Bedrock text model.
Evidence
The final submission package includes the AWS screenshots and external validation results used to verify the architecture, security posture, monitoring, scaling design, and AI integration.
The architecture is designed for a grading/demo workload: small EC2 instances, low-capacity ASG settings, DynamoDB on-demand, capped S3 uploads, Lambda for short event work, and Amazon Nova Micro as a low-cost Bedrock text model.
Bedrock model
Low-cost, not free
Uploads
2 MB max
DynamoDB
On-demand
ASG grading mode
Low capacity
This AWS security architecture review identifies key areas for improvement to enhance security and compliance for the CS 79D final project.
Mode
bedrock
Model
amazon.nova-micro-v1:0
Generated
6/17/2026, 3:00:13 AM
Ensure SSH access is restricted to a specific list of CIDR blocks to minimize the risk of unauthorized access.
Update security group rules to allow SSH traffic only from predefined CIDR blocks.
Establish logging and monitoring for observatory evidence to ensure compliance and detect any anomalies.
Configure CloudWatch Logs and alarms to monitor observatory evidence and set up alerts for any suspicious activity.
Encrypt persistent user storage using AWS KMS to protect sensitive data at rest.
Enable KMS encryption for DynamoDB tables and S3 buckets used for persistent user storage.
Document the restricted SSH CIDR blocks to ensure all team members are aware of the allowed IP ranges.
Create and maintain an up-to-date documentation file listing the CIDR blocks allowed for SSH access.
Evidence