Emerging Market Swap Curve Fitter

FastAPI web service that emits synthetic emerging market swap curves via Server-Sent Events (SSE). The service currently generates randomised curve marks and fits a polynomial over a dense tenor grid to mimic intraday pricing snapshots.

Quick start

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python -m app

The app listens on port 8080 by default. You can also run it with uvicorn app.main:app --reload during development.

Endpoints

• GET / – basic service metadata
• GET /health – readiness probe returning { "status": "ok" }
• GET /curves/stream – SSE stream of curve snapshots. Supports optional interval query parameter (seconds between snapshots).

Consuming the stream

Event data is JSON encoded. A quick way to try the feed is to use curl:

curl -N "http://127.0.0.1:8080/curves/stream?interval=2"

Sample event payload:

{
  "timestamp": "2024-05-11T09:00:00.123456+00:00",
  "tenorYears": [0.5, 1.0, 2.0, 3.0, 4.0, 5.0, 7.0, 10.0, 15.0, 20.0, 30.0],
  "rawRates": [8.32, 8.12, 8.05, 8.11, 8.36, 8.55, 8.87, 9.25, 9.64, 9.85, 10.04],
  "fit": {
    "gridYears": [0.5, 0.75, ...],
    "rates": [8.36, 8.30, ...],
    "polynomialCoefficients": [0.0002, -0.013, 0.25, 0.18, 7.95]
  }
}

Adjust the stream interval or integrate the endpoint into a frontend to visualise the evolving curve.

Continuous delivery

A GitHub Actions workflow at .github/workflows/docker-publish.yml builds the Docker image with Buildx and publishes it to GitHub Container Registry (ghcr.io/<owner>/<repo>) whenever main is updated. Grant the repository's GITHUB_TOKEN the packages: write permission (enabled by default on public repos) to allow the push to succeed.