K3 Analytics
docs

FraudGraph API

Federal program fraud analytics across 278+ datasets. Cross-referenced, name-standardized, and query-ready.

PPP, Medicare, SNAP, federal contracts, SEC filings, EPA enforcement, CFTC actions, provider exclusions, sanctions lists, and more — linked by standardized entity matching across 806 normalized columns.

Why FraudGraph

Name-Standardized Entity Matching

Every entity name and address is normalized across all 278+ tables — lowercase, trimmed, special characters stripped. This means "ACME HEALTH LLC", "Acme Health, L.L.C.", and "acme health llc" all resolve to the same entity. 806 standardized columns enable cross-database joins that would take weeks to build from raw data.

Pre-Built Cross-References

A 41M-row master entity table resolves entities across all 278+ datasets — linking PPP borrowers to sanctions lists, enforcement actions, SEC filings, EPA violations, federal contracts, relief fund payments, and state registrations. Connections that take analysts days to find are already materialized and queryable.

Multi-Program Coverage

Federal contracts (25.9M rows), pandemic relief programs (6.1M PPP + 9M EIDL loans, 420K Provider Relief Fund payments), SBA 7(a)/504 loans (2.1M), Medicare/Medicaid (286M+ claims), SEC EDGAR filings (8.2M), CFPB complaints (14M), EPA ECHO, SNAP retailers, state business registrations (11 states), and global sanctions — all in one queryable warehouse.

AI-Powered Queries

Ask questions in plain English. "Which federal contractors also appear on global sanctions lists?" translates to SQL automatically and runs against the full database. Pro and Enterprise tiers.

Free

$0

  • PPP + LEIE + OFAC tables
  • 25 queries/day
  • 100 rows per query

Pro

$20/mo

  • All 278+ tables
  • 500 queries/day
  • 10K rows per query
  • PPP parquet download

Enterprise

$200/mo

  • All 278+ tables
  • Unlimited queries
  • Unlimited rows
  • AI natural language queries
  • All table downloads

Data Coverage

Lending & Relief

PPP Loans (6.1M)

EIDL Loans & Advances (9M)

SBA 7(a) Loans (1.9M)

SBA 504 Loans (224K)

Provider Relief Fund (420K)

USAspending Grants (9M)

Healthcare

Medicare Claims (59M)

Medicaid Claims (227M)

Part D Prescribers (154M)

NPPES Providers (9.4M)

Open Payments (82M)

SEC & Financial

EDGAR Filings (8.2M)

Insider Transactions (12M)

13F Holdings (87M)

CFTC Enforcement

FDIC Call Reports (4.7K)

Sanctions & Exclusions

OpenSanctions (1.5M)

LEIE (83K)

OFAC SDN (19K)

Consolidated Screening (25K)

World Bank Debarment

Regulatory & Contracts

Federal Contracts (25.9M)

EPA ECHO (3.1M)

FDA Enforcement (69K)

FMCSA Carriers (4.4M)

CFPB Complaints (14M)

Entity Registry & Other

Master Entity Table (41M)

State SOS Filings (11 states)

IRS Exempt Orgs (1.9M)

SNAP Retailers (703K)

HUD Section 8 (24K)

Quick Start: CLI

The fastest way to get started. Works with any AI agent or terminal.

# Install

pip install fraudgraph

# Authenticate

export FRAUDGRAPH_API_KEY=your_key_here

# Check connection

fraudgraph health

# List available tables

fraudgraph catalog

# Run a SQL query

fraudgraph query "SELECT BorrowerName, CurrentApprovalAmount FROM ppp ORDER BY CurrentApprovalAmount DESC LIMIT 10"

# Ask in plain English (Pro/Enterprise)

fraudgraph ai "Which Provider Relief Fund recipients are also excluded from Medicare?"

# Download PPP dataset as Parquet

fraudgraph download ppp

Quick Start: Python

from fraudgraph import FraudGraph

fg = FraudGraph(api_key="your_key_here")

# Run a query

result = fg.query("SELECT * FROM ppp LIMIT 10")

# Ask in plain English

result = fg.ai("Which federal contractors are on sanctions lists?")

# Browse tables

tables = fg.catalog()

Quick Start: HTTP API

Use curl or any HTTP client. All endpoints accept and return JSON.

# Check API status

curl https://api.k3analytics.io/health

# Browse available tables

curl -H "X-API-Key: YOUR_KEY" https://api.k3analytics.io/v1/catalog

# Find PPP borrowers on sanctions lists

curl -X POST -H "X-API-Key: YOUR_KEY" \

-H "Content-Type: application/json" \

-d '{"sql":"SELECT p.BorrowerName, p.CurrentApprovalAmount, o.caption FROM ppp p JOIN opensanctions o ON p.std_org_name = o.std_org_name LIMIT 10"}' \

https://api.k3analytics.io/v1/query

# AI query (Pro/Enterprise)

curl -X POST -H "X-API-Key: YOUR_KEY" \

-H "Content-Type: application/json" \

-d '{"question":"Which Provider Relief Fund recipients are also excluded from Medicare?"}' \

https://api.k3analytics.io/v1/ai/query

K3 Analytics · @K3_Analytics