US Open Data Studio — Gov Data Joins 🇺🇸

Value layer on keyless US federal open data APIs. Join cross-dataset indicators into clean, AI-ready tables — or pull raw series by ID.

Sources: BLS public API v2 (Bureau of Labor Statistics — CPI, unemployment, employment, wages, PPI) and USGS FDSN event API (earthquakes). Both keyless.

Note on data.gov: the US has no national CKAN datastore — data.gov's CKAN API is dead (404) and Census/FRED/BEA/EIA require API keys. This studio uses the keyless federal APIs that actually work.

Licence: US federal data is public domain.

Features

Studio mode (value layer) — $0.02/joined-pull

Six cross-dataset join presets, delivered as CSV / JSON / XLSX with an AI-ready package (JSON Schema + data dictionary + sample rows) and attribution sidecar:

• T1 Inflation & Prices by Year — CPI-U × core CPI × PPI (annual means)
• T2 Labour Market by Year — unemployment rate × employment level × labour force × participation rate
• T3 Wages & Earnings by Year — average hourly × weekly earnings × total nonfarm employment
• T4 Earthquakes by Year — USGS M4.5+ earthquake counts
• T5 Employment by Year — total nonfarm × private-sector employment
• T6 Prices & Wages by Year — CPI-U × average hourly earnings (real-wage view)

Custom joins: bring your own BLS series IDs, key columns, join type (inner/left), aggregation (none/sum/mean), and year params.

Extract mode (raw data pull) — $0.01/extract-pull

Pull 1-5 BLS series raw by ID (e.g. CUUR0000SA0, LNS14000000) or earthquakes, no join, no reshape. Per-dataset files + manifest + attribution.

Input

Field	Type	Default	Notes
mode	select	studio	studio = value layer, extract = raw pull
preset	select	T1	T1-T6 or custom
datasetIds	stringList	—	extract mode: 1-5 series IDs
customDatasets	stringList	—	custom join: 2-6 series IDs
customKey	stringList	[year]	join keys ('year' auto-extracts from BLS year fields)
customJoinType	select	inner	inner / left
customParams	textfield	—	JSON API params (startyear/endyear)
customAgg	select	none	none / sum / mean
format	select	json	json / csv / xlsx
webhookUrl	textfield	—	optional POST target
aiReady	checkbox	true	AI-ready package
includeAttribution	checkbox	true	attribution sidecar

Output

• Dataset: one summary record per run (mode, preset, rows, format, files, year range, run URL) — rendered as a table in the Output tab.
• Key-value store: the joined/extracted file, *_verification.json, *_attribution.json, *_ai_ready.json (studio) or per-dataset files + extract_manifest.json (extract).

Example

# Studio: Inflation & Prices preset, CSV
curl -X POST "https://api.apify.com/v2/acts/<ACTOR_ID>/runs?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"preset":"T1","format":"csv"}'

# Extract: pull CPI + unemployment raw
curl -X POST "https://api.apify.com/v2/acts/<ACTOR_ID>/runs?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"mode":"extract","datasetIds":["CUUR0000SA0","LNS14000000"],"format":"json"}'

Notes

• BLS keyless API: max 25 series + 10 years per request. The engine batches all preset series into one request and loops 10-year chunks for longer ranges.
• BLS keyless has a daily request cap (~25/day per IP) — the actor batches aggressively to stay well under it.
• USGS count endpoint returns plain text (e.g. 7651), not JSON.