Austin Commercial Permit Evidence

Review permit records issued by the City of Austin, selected by the city's published Commercial class, calendar dates, and traceable trade keywords. Each result links back to the source and shows the exact text behind its roof, envelope, HVAC, or renovation classification.

This is an Austin-only research pilot. A permit is evidence of a dated source record. It does not establish that a job is available, that work was completed, that a property currently needs work, or the age of a roof.

Example

{
  "trades": ["roof"],
  "dateField": "issue_date",
  "lookbackDays": 30,
  "maxRecords": 100
}

Choose one or more trades: roof, envelope, hvac, or renovation. The selection uses OR matching. dateField can be issue_date or applieddate. The default window is the 30 calendar days through the run's UTC date. Optional startDate and endDate accept YYYY-MM-DD, inclusive, for a maximum 366-day window. An explicit start date overrides lookbackDays.

The underlying source contains issued permits. Selecting original application dates does not expand it into a database of all submitted or still-pending applications. Source calendar dates are retained as provided; the Actor does not invent their time zone.

Output and buyer workflows

Filter recordType=permit for paid records. Outputs include source permit/project ID, permit number, location, original permit link, original published class, work/permit type, description, selected trade evidence, raw application/issue/status dates, and the current source permit status. Coordinates are included when valid source values exist. Applicant and contractor personal columns are not requested or exported.

tradeEvidence identifies the exact source field, matched text, offsets, and nearby excerpt. Presets use deterministic words, not a model-generated lead score. Roof mentions can refer to rooftop equipment. Renovation mentions can include work outside your niche. Review the source description and permit before qualifying a commercial opportunity.

Three text columns make that evidence easier to review in a CSV export:

Column	What it shows
matchedTradesText	Matched trades in a simple text cell, such as roof, envelope
tradeEvidenceBasis	Source fields supporting each trade, such as roof: description or hvac: permit type only
tradeEvidenceText	Exact matched phrases grouped by trade and source field

The original tradeEvidence entries and offsets remain available. permit type only tells you that an HVAC label is supported by the published Mechanical permit type without a matching description phrase. It is not a confidence score or proof of the particular equipment involved. The text columns add no source requests or event fees.

Useful workflows include weekly supplier territory research, contractor project qualification, and source-backed property work-history research. Save task inputs and use Apify's scheduling/export features for recurring refreshes. Repeated exports use stable source IDs, but a physical project can have several distinct permits.

recordType=query_status and the SUMMARY key-value-store record report exact date/preset scope, source query URL, metadata freshness, filtering, invalid rows, and truncation. complete applies only to this bounded source query. The city's data can change between runs; preserve snapshot/fetch timestamps for comparisons.

Reusable contractor and supplier workflows

Use these input settings to reuse one source adapter across several buyers. Each row uses dateField: "issue_date"; leave explicit start/end dates blank for a rolling window. The cadence is a suggested review frequency, not an already-created schedule.

Workflow	trades	lookbackDays	maxRecords
Weekly roofing supplier review	["roof"]	14	100
Monthly envelope supplier review	["envelope"]	45	100
Weekly HVAC distributor review	["hvac"]	14	100
Weekly renovation contractor review	["renovation"]	14	100
Monthly roof and envelope supplier review	["roof", "envelope"]	45	200
Weekly contractor review across four trades	["roof", "envelope", "hvac", "renovation"]	14	200

Overlapping windows let a reviewer revisit recently issued records as the source updates. They do not guarantee recovery of every delayed or backfilled record. Deduplicate separate exports using recordId in your destination; this Actor does not keep automatic history across runs. A permit returned again in another run is another returned-record event. Within one combined run, a permit matching several trades creates one event.

Start by checking query_status. If it is truncated, shorten the date window or increase the cap within the 500-record limit; a shorter window provides narrower coverage. Review the original class, description, and current permit status before deciding whether a project is relevant. These inputs create source-evidence review queues, not lists of verified open jobs or contractor contact details.

Source scope and maintenance

The source is the City's Issued Construction Permits dataset, via its public API. It advertises daily updates. Dataset metadata's reported update time is included when available; its absence is marked unknown rather than replaced with the Actor's fetch time. The City describes its data as informational and subject to change.

The exact permit_class_mapped=Commercial filter preserves the City's mapping, whose metadata describes non-R permit classes as commercial. This is not independently verified building occupancy and can include multifamily or other source classes. Austin's permitting dataset also includes some extraterritorial areas; not every source location is necessarily inside city limits. The original permit class and jurisdiction are included for your review.

Bounds and charging

One event per distinct valid permit record emitted in the run. Duplicates, rejected source rows, keyword-filter exclusions, failed requests, and status/summary rows are unbilled. Valid capped output remains usable and is labeled truncated.

Maximum 500 output records per run. One data request asks for the requested candidate limit plus one sentinel; precise word matching can return fewer results than the broad server substring predicate. A sentinel flags the source candidate cap even when some candidates are filtered out. No pagination or automatic retries create hidden extra work. One independent bounded metadata request supplies freshness context. Limits are 5 MiB for data, 2 MiB for metadata, and 30/20-second request timeouts.

Use the supplied original sources and City data terms when evaluating reuse. No private lead database, inferred decision maker, automatic contact discovery, outreach, or other-city source is included.

Price and run controls

The launch price is $2 per 1,000 completed units ($0.002 per event), as defined above. Check the current Pricing tab before running. Status and duplicate rows do not add this custom event. Set a maximum charge appropriate for your batch; a run stops when its event budget is exhausted. RUN-SUMMARY records actual accepted events and execution totals.

Automatic replay of an interrupted run is disabled to prevent duplicate charges. Keep the available output, then start a new run for a new execution. A later run is a new billable execution. Completed dataset rows and individually saved files remain available when a budget stops a run. Combined exports, manifests, and snapshots assembled at the end may be absent after a budget stop or interruption; use a sufficient run budget when you need those aggregate files. The initial release uses documented input limits and public source access; external source changes can require maintenance.