Skip to content

Cloud Resume Challenge

Cloud Resume Challenge

Multi-cloud resume website deployed across AWS, Azure, and Google Cloud Platform. Features visitor counters, AI-powered Q&A, and semantic search - all managed with Terraform and CI/CD pipelines.

Project Status

Status: In Progress | Clouds: AWS, Azure, GCP | Goal: Captain-grade (highest tier)

Project Overview

Building a multi-cloud resume website for the Cloud Resume Challenge Bootcamp. The goal is to achieve Captain-grade (highest tier) by deploying to AWS, Azure, and GCP with AI features on AWS. The project demonstrates cloud architecture, infrastructure as code, serverless computing, and AI integration.

Timeline
  • Week 1: Frontend (static site + CDN + DNS)
  • Week 2: Backend (APIs + AI features)
  • Budget: $2-4/month maximum

Architecture

Multi-Cloud Strategy

Content: Shared MkDocs source in docs/ folder

Deployment: Independent per cloud

  • AWS: Full features (AI included)
  • Azure: Basic (visitor counter only)
  • GCP: Basic (visitor counter only)

CI/CD: Separate GitHub Actions workflows per cloud

AWS Architecture (Full Features)

User
  → Route 53 (DNS)
  → CloudFront (CDN)
  → S3 (Static Site)
  → API Gateway
  → Lambda Functions
  → DynamoDB (Visitor Counter)
  → Bedrock/Modal Labs (AI Features)

Azure Architecture (Basic)

User
  → Azure DNS
  → Front Door (CDN)
  → Blob Storage (Static Site)
  → Azure Functions
  → Cosmos DB (Visitor Counter)

GCP Architecture (Basic)

User
  → Cloud DNS
  → Cloud CDN
  → Cloud Storage (Static Site)
  → Cloud Functions
  → Firestore (Visitor Counter)

Key Features

Frontend

  • MkDocs Material Theme: Clean, professional design
  • Responsive Layout: Works on all devices
  • Custom CSS: Portfolio-style aesthetics
  • Project Showcase: Interactive project cards
  • Resume Content: Comprehensive professional history

Backend - Visitor Counter

  • Atomic Increment: NoSQL database operations
  • CORS Support: Cross-origin requests
  • Rate Limiting: 100 requests/IP/minute
  • Multi-Cloud: Separate implementations per cloud

AI Features (AWS Only)

AI Q&A System

  • RAG Pipeline: Retrieval-augmented generation
  • Resume Context: Answers questions about experience
  • Rate Limiting: 5 requests/IP/minute
  • Daily Quota: 100 total requests
  • API Key Authentication: Secure access
  • Pre-computed Embeddings: Build-time generation
  • Cosine Similarity: Fast search results
  • Top 5 Results: Ranked by relevance
  • Rate Limiting: 10 requests/IP/minute

AI Summarizer

  • Build-Time Only: Not per-request
  • Project Summaries: 2-3 sentence descriptions
  • Automatic Injection: Into project pages

Technology Stack

Infrastructure as Code

  • Terraform: All cloud resources
  • Remote State: S3 with DynamoDB locking
  • Modules: Reusable components
  • Variables: Configurable values

Frontend

  • MkDocs: Static site generator
  • Material Theme: Professional design
  • Custom CSS/JS: Enhanced functionality
  • Markdown: Content management

Backend

  • AWS: Lambda, API Gateway, DynamoDB
  • Azure: Azure Functions, Cosmos DB
  • GCP: Cloud Functions, Firestore
  • Python: Lambda/Function code

AI Services

  • Modal Labs: Free credits (primary)
  • Nebius AI: Free credits (backup)
  • AWS Bedrock: Last resort
  • Embeddings: Pre-computed at build time

CI/CD

  • GitHub Actions: Automated deployments
  • Terraform Plan/Apply: Infrastructure updates
  • MkDocs Build: Site generation
  • CloudFront Invalidation: Cache clearing

Security & Cost Optimization

Security

Rate Limiting:

  • AI Q&A: 5 requests/IP/minute
  • AI Search: 10 requests/IP/minute
  • Visitor Counter: 100 requests/IP/minute

Quotas:

  • Daily AI request limit: 100 total
  • CloudWatch alarms on unusual usage

Authentication:

  • API keys for AI endpoints (Secrets Manager)
  • Request size limits (500 chars for questions)
  • Input validation and sanitization

Cost Optimization

Budget: $2-4/month maximum

AI Services:

  • Use Modal Labs free credits first
  • Use Nebius AI free credits as backup
  • AWS Bedrock only as last resort
  • AI Search uses pre-computed embeddings (free)
  • AI Summarizer runs at build-time only

Free Tier Usage:

  • AWS: S3, CloudFront (12 months), Lambda, DynamoDB
  • Azure: Blob Storage, Functions, Cosmos DB
  • GCP: Cloud Storage, Cloud Functions, Firestore

Skills Demonstrated

Cloud Architecture: Multi-cloud deployment, CDN configuration, DNS management, serverless computing

Infrastructure as Code: Terraform modules, state management, resource tagging, best practices

Serverless Development: Lambda functions, API Gateway, Azure Functions, Cloud Functions

AI/ML Integration: RAG pipelines, embeddings, semantic search, LLM integration

DevOps: CI/CD pipelines, GitHub Actions, automated deployments, monitoring

Security: Rate limiting, API authentication, input validation, secrets management

Cost Optimization: Free tier usage, resource tagging, budget monitoring, efficient architecture

Frontend Development: MkDocs, Material theme, custom CSS/JS, responsive design

Deployment Checklist

Before deploying to production:

  • Run terraform plan and review changes
  • Check cost estimate (should be < $5/month)
  • Verify rate limiting is configured
  • Test visitor counter works
  • Test AI features (if AWS)
  • Verify HTTPS works
  • Check DNS resolution
  • Run smoke tests
  • Set up CloudWatch alarms

← Back to Projects