AWS Serverless Resume

A serverless resume website that runs the same FastAPI code locally (Docker) and in production (AWS Lambda). Includes a floating AI chat widget powered by Bot Factory. Update your resume via Excel, deploy with Terraform.

Live: robrose.info

---

Quick Start

1. Clone and Configure

git clone https://github.com/mr-flowjangles/aws-serverless-resume.git
cd aws-serverless-resume
cp .env.example .env
mkdir -p _scratch

Edit .env with your reCAPTCHA secret key and SES email addresses. See Configuration below for details.

2. Add Your Resume Data

Open scripts/resume-data-template.xlsx in Excel or Google Sheets and fill in your profile, work experience, education, and skills.

3. Run Locally

make up

Open http://localhost:8080. API docs at http://localhost:8080/api/docs.

4. Deploy to AWS

See README_DEPLOY.md for complete deployment instructions.

---

How It Works

Local Development (Docker)

Browser → Nginx → FastAPI (uvicorn) → LocalStack DynamoDB

Production (AWS)

Browser → CloudFront → Lambda Function URL (FastAPI + Mangum) → DynamoDB
       → Chat Widget → Bot Factory Lambda (SSE streaming) → Bedrock Claude

Same Python code, different wrapper. Mangum adapts FastAPI to Lambda's event format. The RobbAI chat widget connects client-side directly to Bot Factory's streaming endpoint.

---

Project Structure

aws-serverless-resume/
├── api/                        # Backend (runs in Lambda)
│   ├── handlers/               # Business logic (environment-agnostic)
│   │   ├── contact.py          # Contact form + reCAPTCHA + SES
│   │   ├── db.py               # DynamoDB connection
│   │   ├── health.py           # Health check
│   │   └── resume_all.py       # Resume data (cached)
│   ├── routers/                # FastAPI route definitions
│   ├── tests/                  # pytest suite
│   ├── main.py                 # FastAPI app setup
│   ├── lambda_handler.py       # Mangum wrapper (Lambda entry point)
│   ├── seed.py                 # Auto-seeds DynamoDB locally
│   ├── requirements.txt        # Full dependencies (local dev)
│   ├── requirements-lambda.txt # Slim dependencies (Lambda only)
│   ├── Dockerfile              # Local dev container
│   └── Dockerfile.lambda       # Lambda build container
├── app/                        # Frontend (served by CloudFront)
│   ├── index.html              # Single-page app (Welcome, About, Projects, etc.)
│   ├── scripts/
│   │   ├── api.js              # API client
│   │   ├── navigation.js       # Tab navigation and section loading
│   │   ├── architecture.js     # Architecture diagrams
│   │   ├── projects.config.js  # Project card definitions
│   │   ├── loaders.js          # Section data loaders
│   │   ├── contact.js          # Contact form handler
│   │   └── chat-widget.js      # RobbAI floating chat (connects to Bot Factory)
│   └── assets/
│       ├── main.css            # Core styles
│       ├── mobile.css          # Mobile responsive styles
│       ├── architecture.css    # Architecture section styles
│       └── chat-widget.css     # Chat widget styles
├── scripts/
│   ├── resume-data-template.xlsx  # Resume data (single source of truth)
│   ├── load_resume.py          # Excel → DynamoDB loader
│   ├── build-lambda.sh         # Lambda package builder
│   └── init-dynamodb.sh        # LocalStack table setup
├── terraform/                  # Infrastructure as Code
├── docker-compose.yml          # Local development setup
├── Makefile                    # Convenience commands
└── README_DEPLOY.md            # Deployment guide

---

Configuration

Google reCAPTCHA (Required for Contact Form)

1. Get keys at https://www.google.com/recaptcha/admin (v2, "I'm not a robot")
2. Add domains: localhost (dev) and your production domain
3. Put the secret key in .env as RECAPTCHA_SECRET_KEY
4. Put the site key in app/index.html in the data-sitekey attribute

AWS SES (Required for Email Delivery)

aws ses verify-email-identity --email-address your@email.com --region us-east-1

Add to .env:

SES_FROM_EMAIL=your@email.com
SES_TO_EMAIL=your@email.com

AWS Credentials

aws configure

Domain Name (Optional)

Register in Route 53 and update terraform/variables.tf. Terraform handles SSL and DNS.

---

Update Resume Data

# Local
docker compose exec api python /app/scripts/load_resume.py /app/scripts/resume-data-template.xlsx

# Production
AWS_ENDPOINT_URL="" AWS_REGION="us-east-1" python3 scripts/load_resume.py path/to/your-resume-data.xlsx

---

Tech Stack

Backend: Python 3.12, FastAPI, Mangum, boto3 Database: DynamoDB (LocalStack locally, AWS in production) Frontend: HTML, CSS, vanilla JavaScript AI Chat: RobbAI widget powered by Bot Factory (SSE streaming) Email: AWS SES Security: Google reCAPTCHA v2 Local Dev: Docker, Docker Compose, Nginx, LocalStack AWS: Lambda (Function URL + streaming), API Gateway, S3, CloudFront, Route 53, ACM, DynamoDB, SES Infrastructure: Terraform (100% IaC) Testing: pytest, pre-commit hooks

---

Running Tests

docker compose exec api pytest tests/ -v

Pre-commit hooks run tests automatically before every commit:

pre-commit install

---

Cost Breakdown

For a personal resume site (<1,000 visitors/month): ~$3-10/month. Route 53 ($0.50), CloudFront ($1-2), Lambda/API Gateway/DynamoDB/SES (free tier), S3 ($0.50), ACM (free).

---

License

© 2026 Rob Rose. All rights reserved.

This project is provided for personal and educational use. If you fork or reuse, please provide attribution and link back to the original repository.

---

Built with FastAPI, Mangum, Terraform, LocalStack, and Claude as a development partner.