Data API Overview

The NewBizBot Data API provides access to public government datasets through a standards-based REST interface powered by PostgREST.

Available Datasets

Dataset	Table	Records	Description
PPP Loans	`ppp_loans`	~11.5M	SBA Paycheck Protection Program loan records (FOIA)
FMCSA Census	`fmcsa_census`	~1.96M	FMCSA Motor Carrier Census — active motor carrier records

Key Features

PostgREST query language — filter, sort, paginate, and search using URL query parameters
Full-text search — search by name with PostgreSQL tsvector
Aggregate RPCs — pre-built aggregate functions for stats by state or industry
API key authentication — Clerk-powered API keys for machine-to-machine access
Rate limiting — 100 requests/minute per API key
CORS enabled — use directly from browser-based applications

Quick Examples

# Find restaurants (NAICS 7225*) in Texas with loan amount > $50K
curl "https://newbizbot.ai/api/data/ppp_loans?borrower_state=eq.TX&naics_code=like.7225*&initial_approval_amount=gt.50000&limit=5" \
  -H "Authorization: Bearer YOUR_API_KEY"

# Find motor carriers in California with 50+ power units
curl "https://newbizbot.ai/api/data/fmcsa_census?phy_state=eq.CA&power_units=gte.50&limit=5" \
  -H "Authorization: Bearer YOUR_API_KEY"

Getting Started

Create an API key from the dashboard
Read the Authentication guide to learn how to authenticate requests
Browse the query syntax reference below for filtering, sorting, and pagination
Browse the NAICS Code Reference to understand industry codes (PPP Loans)
See the PPP Loans or FMCSA Census dataset pages for schemas, querying examples, and aggregate functions

Query Syntax Reference

The Data API uses PostgREST query syntax. All filtering, sorting, and pagination is done via URL query parameters.

Base URL

https://newbizbot.ai/api/data/{table}

Replace {table} with ppp_loans or fmcsa_census. See each dataset's page for available columns and indexes.

Filtering

Add filters as query parameters using the format column=operator.value:

# Exact match
?borrower_state=eq.TX

# Greater than
?initial_approval_amount=gt.100000

# Pattern matching (wildcards)
?naics_code=like.7225*

# Multiple filters (AND)
?borrower_state=eq.TX&naics_code=like.7225*&initial_approval_amount=gt.50000

Available Operators

Operator	Description	Example
`eq`	Equals	`?borrower_state=eq.TX`
`neq`	Not equals	`?loan_status=neq.Exemption 4`
`gt`	Greater than	`?initial_approval_amount=gt.100000`
`gte`	Greater than or equal	`?jobs_reported=gte.10`
`lt`	Less than	`?initial_approval_amount=lt.50000`
`lte`	Less than or equal	`?term=lte.24`
`like`	Pattern match (case-sensitive)	`?naics_code=like.7225*`
`ilike`	Pattern match (case-insensitive)	`?borrower_name=ilike.restaurant`
`is`	Null check	`?forgiveness_amount=is.null`
`in`	In list	`?borrower_state=in.(TX,CA,NY)`
`cs`	Array contains (for array columns)	`?cargo_types=cs.{"General Freight"}`

Full-Text Search

Search by name using PostgreSQL full-text search:

# PPP Loans — search borrower name
?borrower_name_search=fts.mcdonalds

# FMCSA Census — search carrier name
?legal_name_search=fts.werner

Full-text search columns use GIN indexes and are optimized for fast lookups across millions of records.

Selecting Columns

By default, a core subset of columns is returned. Request specific columns with select:

?select=borrower_name,borrower_city,initial_approval_amount,naics_code

Use select=* to get all columns.

Sorting

# Sort by a column descending
?order=initial_approval_amount.desc

# Sort by multiple columns
?order=borrower_state.asc,initial_approval_amount.desc

Only indexed columns can be used for sorting. See each dataset page for the list of sortable columns.

Pagination

Limit & Offset

# First 50 results
?limit=50

# Next 50 results
?limit=50&offset=50

Limits:

Maximum limit: 1,000 (default if omitted)
Maximum offset: 10,000

For deeper pagination, use keyset pagination:

# First page
?order=id&limit=1000

# Next page (use last id from previous response)
?id=gt.12345&order=id&limit=1000

Range Header

Alternatively, use the Range header:

curl "https://newbizbot.ai/api/data/ppp_loans?borrower_state=eq.FL" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Range: 0-49"

Counting

To get a count of matching rows without fetching data:

curl -I "https://newbizbot.ai/api/data/ppp_loans?borrower_state=eq.CA&limit=0" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Prefer: count=exact"

The total count appears in the Content-Range response header: Content-Range: */48523

Aggregate RPCs

Both datasets provide pre-built RPC endpoints for aggregate queries. Call them at:

https://newbizbot.ai/api/data/rpc/{function_name}

See each dataset page for available RPC functions, parameters, and examples:

Query Guardrails

To ensure fast responses, these guardrails are enforced:

Rule	Details
Filter required	At least one filter must use an indexed column (see each dataset's Fast-Lookup Columns)
Sortable columns	ORDER BY is restricted to indexed columns
Limit cap	Maximum 1,000 rows per request
Offset cap	Maximum 10,000 offset (use keyset pagination beyond this)
OR clause limit	Maximum 3 clauses in an `or` expression

Guardrail violations return a 400 response with an error and message explaining what to fix.

On this page