AI Platform Capstone

A runnable FastAPI capstone that brings together the major AI foundation/tooling patterns demonstrated across this portfolio: document retrieval, model routing, feature flags, response evaluation, observability, and CI/CD quality gates.

Architecture

Client Request → Feature Flag Check → Document Retrieval → Model Routing → Synthetic Response Generation → Response Evaluation → Request Trace / Observability → API Response

The project intentionally uses synthetic models and local retrieval so it runs without proprietary API keys.

Engineering Skills Demonstrated

• Python
• FastAPI
• Pydantic
• REST API design
• RAG-style retrieval
• Text chunking and similarity ranking
• Model routing and fallback
• Feature flags
• AI response evaluation
• Request tracing and observability
• Health checks
• Automated testing
• GitHub Actions CI
• Docker
• Modular service architecture

Endpoints

• GET /health
• GET /v1/models
• GET /v1/flags
• POST /v1/documents
• POST /v1/query
• GET /v1/metrics

Run on macOS

python3 -m venv .venv
source .venv/bin/activate
python3 -m pip install -e ".[dev]"
uvicorn app.main:app --reload

Open /docs for interactive FastAPI documentation.

Tests

pytest -q

Docker

docker build -t ai-platform-capstone .
docker run -p 8000:8000 ai-platform-capstone

CI/CD

.github/workflows/ci.yml runs unit/API tests and a quality-gate test on pushes and pull requests.

Portfolio Note

This is a synthetic engineering portfolio project. It demonstrates architecture and tooling patterns; it does not claim production deployment or professional AI-engineering experience.