OFAC Sanctions API Demo

A TypeScript serverless API that exposes U.S. Treasury OFAC SDN data through a documented OpenAPI interface with Swagger UI, ReDoc, fuzzy search, entity lookup, metadata, scheduled data refresh, and refresh-delta metadata.

Built as the compliance data layer for a larger financial intelligence prototype.

GET /api/search?q=maduro
GET /api/entity/{uid}
GET /api/meta

Try in Swagger UI · Read ReDoc · View GitHub

What this demonstrates

TypeScript API design

The API is implemented with Netlify Functions and a small, explicit REST surface for search, entity lookup, metadata, and controlled dataset refresh. It demonstrates a practical Serverless architecture for public compliance data.

OpenAPI documentation

The project includes an OpenAPI specification, Swagger UI, ReDoc, and practical cURL, JavaScript, and Python examples.

Compliance data integration

The data pipeline fetches OFAC Sanctions List Service CSV exports, parses SDN, alternate-name, and address files, and stores a compressed searchable dataset in Netlify Blobs.

Fuzzy search

Fuse.js powers approximate matching across entity names and aliases, which is useful for real screening workflows where names are often incomplete or spelled differently.

Production-style API patterns

The project demonstrates validation, structured JSON responses, cache headers, CORS handling, scheduled refresh design, update-secret authentication, and delta summaries between refreshes.

Financial intelligence context

This API began as the compliance data layer for a broader financial intelligence prototype: combine sanctions data, country-level signals, world news, AI-assisted sentiment analysis, and uncertainty modeling into a workflow for geopolitical risk research.

The companion concept is ED 209, a Python FastAPI prototype that explores uncertainty-aware sanctions screening with Subjective Logic. Instead of treating a fuzzy match as a crude yes/no answer, the risk engine can explain which evidence is strong, which evidence is missing, and whether the next action should be clear, escalate, block, or gather more information.

Quick Start

Search for an entity:

curl "https://sdn-openapi.netlify.app/api/search?q=Maduro&limit=5"

Example response shape:

{
  "ok": true,
  "q": "Maduro",
  "count": 2,
  "results": [
    {
      "uid": "22790",
      "name": "MADURO MOROS, Nicolas",
      "type": "individual",
      "score": 0.06,
      "programs": ["VENEZUELA"]
    }
  ]
}

Check dataset metadata:

curl "https://sdn-openapi.netlify.app/api/meta"

Data Provenance and Freshness

All source data comes from the U.S. Treasury Office of Foreign Assets Control (OFAC) Sanctions List Service exports.

Official OFAC source: OFAC SDN Human-Readable Lists

Automatic refresh: scheduled Netlify Function
Manual refresh: POST /api/update, protected by a configured update secret
Source files: SDN, ALT, and ADD CSV exports
Verification: /api/meta exposes source URLs, row counts, SHA-256 hashes, and the latest refresh delta summary

API endpoints

`GET /api/search?q=query`

Search OFAC-designated entities by name or alias using fuzzy matching.

`GET /api/entity/{uid}`

Retrieve the stored entity record for a specific OFAC entity UID.

`GET /api/meta`

Return dataset metadata, including source URLs, counts, hashes, refresh timestamp, and delta summary.

`POST /api/update`

Refresh the dataset from OFAC source files. Manual calls require the configured update secret.

Why this matters

Financial and compliance systems often need to reason about messy entity data. Names are misspelled, aliases differ, countries matter, and a fuzzy score alone does not explain whether a match should be blocked, ignored, or escalated.

This project focuses on the API layer: making public sanctions data easy to query, document, verify, and integrate. The larger portfolio concept builds on this layer with uncertainty-aware screening and AI-assisted risk analysis.

Stack

API: TypeScript, Netlify Functions
Storage: Netlify Blobs
Search: Fuse.js
Data source: U.S. Treasury OFAC SLS exports
Docs: OpenAPI, Swagger UI, ReDoc
Testing: Vitest

Built by Robert Melrose as a fintech engineering portfolio project.