PulseData | E-Commerce Analytics Platform

PulseData is a production-style backend analytics project built with .NET 8, PostgreSQL, and a practical ETL pipeline.

This repository was designed to feel like real engineering work: ingest data, model it well, expose useful analytics, and keep operations healthy with tests and CI.

What You Get

• Clean architecture split: Core, Infrastructure, API, ETL
• SQL-first analytics with Dapper (performance and clarity)
• Liveness/readiness health endpoints for orchestration
• Docker-based local stack
• Automated unit and integration tests for health behavior
• CI pipeline on push and pull request to main

5-Minute Start

git clone https://github.com/kill74/PulseData.git
cd PulseData
# Start PostgreSQL + pgAdmin
docker-compose up -d
# Run ETL (loads sample orders)cd src/PulseData.ETL
dotnet run
# Run APIcd ../PulseData.API
dotnet run

Expected services:

• PostgreSQL: localhost:5432
• pgAdmin: http://localhost:8080
• API (local): https://localhost:5001 or http://localhost:5000

Full Docker Mode

docker-compose --profile with-api up -d

Expected services:

• PostgreSQL: localhost:5432
• pgAdmin: http://localhost:8080
• API (container): http://localhost:8000

Verify It Works

# Liveness
curl http://localhost:5000/api/health/live
# Readiness
curl http://localhost:5000/api/health/ready

Readiness semantics:

• 200 when dependencies are healthy
• 408 when a dependency check times out
• 503 when a required dependency is unavailable

Note: health endpoints are anonymous by design to support probes from load balancers and orchestrators.

API Surface

Health

Method	Endpoint	Purpose
GET	/api/health/live	Process liveness
GET	/api/health/ready	Dependency readiness

Analytics

Method	Endpoint
GET	/api/analytics/sales-summary
GET	/api/analytics/top-products?limit=10
GET	/api/analytics/customer-stats?limit=20
GET	/api/analytics/category-performance
GET	/api/analytics/sales-by-period?startDate=2026-01-01&endDate=2026-01-31

Orders

Method	Endpoint
GET	/api/orders?page=1&pageSize=20&status=shipped
GET	/api/orders/{id}
GET	/api/orders/by-customer/{customerId}

Customers

Method	Endpoint
GET	/api/customers?page=1&pageSize=20
GET	/api/customers/{id}

Products

Method	Endpoint
GET	/api/products?page=1&pageSize=20&activeOnly=true
GET	/api/products/{id}
GET	/api/products/by-category/{categoryId}

Testing

Run the API test suite:

dotnet test ./tests/PulseData.API.Tests/PulseData.API.Tests.csproj

Current coverage includes:

• Unit behavior tests for HealthController
• Integration HTTP tests using WebApplicationFactory:
  • /api/health/live
  • /api/health/ready timeout path (408)
  • /api/health/ready dependency failure path (503)

CI

Workflow: .github/workflows/ci.yml

CI validates on push and pull request to main:

• API restore + build
• Test project restore + build
• Health endpoint test execution

Repository Layout

PulseData/
src/
PulseData.Core/
PulseData.Infrastructure/
PulseData.API/
PulseData.ETL/
sql/
tests/
PulseData.API.Tests/
.github/workflows/
docker-compose.yml
Makefile

Useful Commands

# Reliable API build
dotnet build ./src/PulseData.API/PulseData.API.csproj
# Run health-related tests
dotnet test ./tests/PulseData.API.Tests/PulseData.API.Tests.csproj
# Start infra only
docker-compose up -d
# Start full stack
docker-compose --profile with-api up -d
# API logs (container mode)
docker-compose logs -f api

Documentation

• QUICK_START.md
• DEVELOPMENT.md
• DOCKER.md
• docs/architecture.md

License

MIT