US Federal Court Corporate Parties & Law Firm Finder

Find corporate parties and explicitly recorded law firms in published U.S. federal court opinions on GovInfo. The Actor produces one structured record per unique case package, with corporate party names, Nature of Suit, opinion dates, source links, and page references for supported firm relationships. Download JSON for detailed evidence or CSV for a research table, Clay, or an n8n workflow.

The source is GovInfo's USCOURTS collection. Access uses public sitemaps, MODS metadata, and opinion PDFs. No GovInfo API key, source account, proxy, or LLM is required.

Automatic firm extraction currently supports recognized Federal Circuit and Sixth Circuit counsel layouts with an unambiguous singular caption party. The recorded firm name must have a legal-form marker, such as LLP, or an explicit law-firm designation. Other courts, plural or ambiguous caption groups, and unsupported layouts can produce corporate-case metadata with empty firm arrays. An empty firm list does not establish that a party has no counsel.

Price: $0.006 per unique GovInfo case package returned — $6 per 1,000 result rows. One case remains one billed row when several matching opinions are inspected. A valid corporate-party case record is billable even when its firm list is empty; check the extraction status and evidence fields before treating a row as a company-to-firm match.

What a result means

Each docket_id is a GovInfo case-package ID such as USCOURTS-ca13-23-01048. GovInfo groups published opinions for the case under this package; each individual opinion has a granule ID. The record combines metadata with explicit counsel evidence from the matching opinions selected for inspection.

The Actor selects up to maxOpinionsPerDocket matching opinions per case, with a default of three. It attempts to read the first three PDF pages of each selected opinion. Within the supported Federal and Sixth Circuit layouts, a firm link requires an unambiguous singular caption role and a firm name with a legal-form marker or explicit law-firm designation. The output records the opinion, PDF URL, and one-based PDF page for each supported link. Ambiguous suffixless names and group-caption associations are omitted.

Corporate names are selected using legal-form markers recorded in the source party names. This is a practical filter for corporate parties, not company-registry verification. A metadata role such as Appellee is preserved as recorded. A more specific role such as Defendant-Appellee is supplied only when supported by the opinion text.

representation_current_status is always not_asserted. The observations describe the cited documents and do not establish that any law firm is currently retained. filing_date is always null: the workflow supplies opinion dates and does not establish the date the underlying case was filed.

Quick start

Use the default sitemap input to explore Patent opinions from the Federal Circuit in January 2024:

{
  "sourceMode": "sitemap",
  "court": "ca13",
  "nature_of_suit": "830",
  "sitemapYears": [2024],
  "opinion_date_from": "2024-01-01",
  "opinion_date_to": "2024-01-31",
  "maxResults": 3,
  "maxCostUsd": 0.018,
  "maxSourceRequests": 60,
  "maxRunSeconds": 240,
  "requestIntervalSeconds": 1,
  "maxOpinionsPerDocket": 3,
  "maxPdfBytes": 8388608
}

Open the run's Case results output to review records. Use Download JSON to preserve arrays and detailed representation links, or Download CSV for the scalar text companions. The result limit is a ceiling; the source filters and processing budgets can produce fewer than three rows.

Inspect known cases directly

Choose packages when you already have GovInfo case-package IDs. This avoids sitemap discovery and is useful for inspecting a specific case or a small research list:

{
  "sourceMode": "packages",
  "court": "",
  "nature_of_suit": "",
  "opinion_date_from": "2024-01-01",
  "opinion_date_to": "2024-12-31",
  "packageIds": [
    "USCOURTS-ca13-23-01048",
    "USCOURTS-ca6-23-05456"
  ],
  "maxResults": 2,
  "maxCostUsd": 0.012,
  "maxOpinionsPerDocket": 3
}

An empty court accepts supported federal courts in packages mode. An empty nature_of_suit accepts any Nature of Suit. Opinion-date filters still apply. Supply case-package IDs, not PDF URLs, individual opinion IDs, or identifiers from another source.

Discovery and coverage

In sitemap mode, the Actor visits the requested court and publication-year partitions, follows their case-package entries, and checks the source metadata. sitemapYears controls GovInfo publication-year discovery. It does not select case filing years, and it is not a guarantee that every opinion dated in the requested range appears in those partitions. The opinion date, court, and Nature of Suit are checked again against package metadata before a case is returned.

Discovery and document inspection are bounded by result, charge, request, runtime, opinion-count, and PDF-size limits. A run can finish after producing useful partial results. Review RUN_SUMMARY to understand the stopping condition, and compare matching_opinions_available, source_opinion_ids, and opinions_inspected to assess each case's inspection coverage.

GovInfo describes USCOURTS as opinions from selected federal appellate, district, bankruptcy, and certain national courts, generally from 2004 onward, with gaps and some earlier material. It is not a complete docket database. Missing opinions, unavailable party metadata, image-only or unreadable PDFs, and counsel information outside the inspected pages can all reduce coverage. Corporate-party metadata coverage is broader than automatic firm extraction: firm links currently require the supported Federal or Sixth Circuit singular-caption layouts and explicit firm markers. A valid corporate-case record can still be returned and billed when its PDF is unavailable or its counsel layout is unsupported. Source: GovInfo USCOURTS collection help.

The Actor uses documented sitemap and content access rather than GovInfo's API. See GovInfo sitemaps and the GovInfo Developer Hub for the source's public data-access methods.

Input reference

Field	Default	Accepted values and meaning
sourceMode	"sitemap"	"sitemap" discovers court/year partitions; "packages" inspects the supplied case-package IDs.
court	"ca13"	GovInfo court code. Required and nonblank in sitemap mode. Empty is allowed in packages mode. A supplied value must match source metadata.
nature_of_suit	"830"	Three-digit federal Nature of Suit code; 830 means Patent and 840 means Trademark. Empty accepts any Nature of Suit. The code is checked through metadata and the Actor's code-to-label mapping.
sitemapYears	[2024]	One to three integer publication-year partitions for sitemap discovery. Packages mode uses packageIds instead.
opinion_date_from	"2024-01-01"	Inclusive earliest opinion date, as a valid YYYY-MM-DD calendar date.
opinion_date_to	"2024-01-31"	Inclusive latest opinion date. Must be on or after the beginning date.
packageIds	[]	Nonempty array of GovInfo case-package IDs in packages mode. Duplicate IDs do not create additional rows within the run.
maxResults	3	Integer from 1 to 1,000. Maximum unique case-package rows.
maxCostUsd	0.018	Number from 0 to 100. Application ceiling for result charges at $0.006 per case. Zero permits no output rows.
maxSourceRequests	60	Integer from 1 to 3,000. Maximum source requests, including discovery, metadata, PDF downloads, and retries.
maxRunSeconds	240	Integer from 30 to 3,600. Maximum processing time, including request spacing and backoff.
requestIntervalSeconds	1	Number from 1 to 60. Minimum interval between source requests. A longer source-requested retry delay takes precedence.
maxOpinionsPerDocket	3	Integer from 1 to 10. Maximum matching opinions selected per case. Only the first three PDF pages of each selected opinion are inspected.
maxPdfBytes	8388608	Integer from 1,048,576 to 16,777,216 bytes, or 1–16 MiB. Maximum size of each downloaded PDF.

The form supplies these defaults. In JSON input, use an empty string explicitly to clear the optional court or Nature of Suit filter; omitting the field retains its default.

Output reference

Case and source fields

Field	Type	Meaning
docket_id	string	Unique GovInfo case-package ID and within-run billing key.
docket_number	string	Source case number.
case_name	string	Case caption from GovInfo metadata.
filing_date	null	Always null; the case filing date is not established.
court_id	string	GovInfo federal court code.
nature_of_suit	string or null	Resolved Nature of Suit code, or null when unavailable.
nature_of_suit_label	string or null	Source Nature of Suit label, or null when unavailable.
source	string	Always govinfo_uscourts.
record_scope	string	Always case_with_published_opinions.
source_docket_url	string	Official GovInfo case-details URL.
observed_at	string	UTC observation timestamp in ISO 8601 format.

Opinion coverage and firm evidence

Field	Type	Meaning
corporate_party_names	array of strings	Corporate party names supported by source metadata and the legal-form filter.
attorney_firm_names	array of strings	Firms with explicit legal-form markers or law-firm designations, linked through recognized Federal/Sixth Circuit singular-caption layouts. Other courts and unsupported or ambiguous layouts can yield an empty array.
opinion_dates	array of strings	Unique sorted dates of selected matching opinions. These dates remain present when a selected PDF cannot be read.
source_opinion_ids	array of strings	Selected GovInfo opinion granule IDs, limited by maxOpinionsPerDocket.
source_opinion_urls	array of strings	Official opinion-details URLs corresponding to the selected IDs.
matching_opinions_available	integer	Total metadata-verified opinions in the package that match the date filters, before the per-case selection limit.
opinions_inspected	integer	Selected opinions whose first three PDF pages were successfully parsed.
corporate_party_roles	array of objects	Each object preserves party_name and metadata_party_role.
recorded_representation_links	array of objects	Supported singular-party/firm links from recognized Federal/Sixth Circuit counsel layouts, with opinion and page provenance. Unsupported courts/layouts and ambiguous or plural groups yield no links.
firm_extraction_status	string	Inspection and matching outcome, from the status table below.
representation_current_status	string	Always not_asserted.

Every representation-link object contains party_name, metadata_party_role, recorded_party_role, attorney_firm_names, source_opinion_id, source_document_url, evidence_page, evidence_type, and representation_current_status. evidence_page uses one-based PDF pages. Metadata and document roles are preserved separately so an appellate role is not silently converted into a trial-court role.

The flat firm list is convenient for case-level research. Use recorded_representation_links when assigning a particular firm to a particular corporation. A case can have several corporate parties on opposing sides; joining every corporate name to every firm in the flat lists would invent relationships.

firm_extraction_status	Interpretation
firms_found	All selected PDFs were inspected and at least one explicit representation link was found.
no_explicit_firm_match	All selected PDFs were read, but no supported link was found. This can mean an unsupported court or layout, an ambiguous or plural caption, or no qualifying firm name; it does not establish that the source lacks counsel information.
partial_opinion_inspection	Some selected PDFs could not be read, regardless of whether other selected opinions supplied firm evidence.
pdf_unavailable_or_unsupported	None of the selected PDFs could be read.

Even firms_found describes the selected opinions and inspected pages, not complete case history. opinion_dates is a unique date list and should not be paired positionally with source_opinion_ids; several opinions may share a date.

CSV companions

corporate_party_names_text, attorney_firm_names_text, and opinion_dates_text join their corresponding arrays with ; . They make CSV rows easier to import without flattening nested evidence. The JSON export retains the full arrays and representation-link objects.

Examples from public opinions

These public source examples illustrate the firm evidence covered by the supported layouts. The page references below count from the first PDF page.

Opinion	Corporation and recorded role	Firm covered by automatic extraction	Source
Plotagraph, Inc. v. Lightricks, Ltd.; Federal Circuit, January 22, 2024	Lightricks, Ltd.; Defendant-Appellee in the caption	Norton Rose Fulbright US LLP is associated with the defendant-appellee in the counsel paragraph.	Official opinion, page 1
FedEx Ground Package System, Inc. v. Route Consultant, Inc.; Sixth Circuit, April 1, 2024	Route Consultant, Inc.; Defendant-Appellee in the caption	BURR & FORMAN LLP appears for the appellee in the counsel section.	Official opinion, page 1

The Plotagraph plaintiff side is a plural caption group. Its corporate-party metadata can be retained, but automatic extraction omits that group's firm associations. The singular Lightricks defendant association shown above is within the supported scope.

The FedEx opinion also names LEWIS THOMASON for the appellee. Version 0.2 omits ambiguous firm names without a legal-form marker or explicit law-firm designation, so this name is not included in the automatic firm list.

The corresponding case packages are USCOURTS-ca13-23-01048 and USCOURTS-ca6-23-05456. The packages-mode example above selects both. The opinions document historical representation; the Actor does not assert a present relationship.

Pricing and budget behavior

The billing unit is one unique GovInfo case package saved as one dataset row. At $0.006 per row, the result-event totals are:

Returned unique cases	Result-event charge
1	$0.006
3	$0.018
100	$0.60
1,000	$6.00

The Actor uses Apify's single automatic default-dataset-item event. There is no Actor-start charge or separate charge for each selected opinion. Operational summaries and source errors are stored outside the result dataset and do not create paid result rows. See Apify's pay-per-event documentation for platform event behavior.

maxCostUsd controls the application's result allowance. Apify's maximum total charge is a separate run setting, available through maxTotalChargeUsd when using its API. The lower applicable allowance limits the number of rows. A budget that cannot afford the next whole $0.006 row produces no further row; for example, $0.010 permits at most one result. Setting a larger result budget does not expand the request, time, or source-coverage limits automatically.

Deduplication is scoped to one run. Starting a new run can return and bill a case that appeared in an earlier run. The key-value store contains CHECKPOINT for internal run progress and RUN_SUMMARY for processing counts and stopping information. These records are separate from the downloadable case dataset.

JSON and CSV exports

The output tab supplies Download JSON, Download CSV, and Run summary links. For direct integration, obtain the run's defaultDatasetId and use Apify's official dataset endpoint:

GET https://api.apify.com/v2/datasets/DATASET_ID/items?format=json&clean=true

For a compact CSV that preserves the intended case-level columns:

GET https://api.apify.com/v2/datasets/DATASET_ID/items?format=csv&clean=true&attachment=true&fields=docket_id,case_name,court_id,nature_of_suit,opinion_dates_text,corporate_party_names_text,attorney_firm_names_text,matching_opinions_available,opinions_inspected,firm_extraction_status,source_docket_url

Dataset access follows your Apify storage permissions. Use your Apify credential through an Authorization: Bearer ... header when required; GovInfo source credentials are not part of this workflow. Apify supports JSON/CSV export, field selection, and pagination with offset and limit. See Get dataset items.

Clay

Download CSV and import it into a new or existing Clay table using Clay's CSV source. Preserve docket_id as the case identifier and map the _text fields as text columns. Keep the source URL and extraction status visible for review. If you need one corporation per row, expand the structured JSON carefully and use the representation links for any company-to-firm assignments. See Clay's CSV import guide.

n8n

After the Actor run finishes, use an HTTP Request node with method GET and the dataset JSON URL above. Select a JSON response and configure an Apify credential if the dataset requires authorization. Pass the returned records into your downstream workflow, retaining docket_id, the source URLs, and the evidence objects. For a CSV file workflow, request format=csv instead. See n8n HTTP Request documentation.

Source use

GovInfo provides public access to these court opinions and metadata. Its public-domain and copyright policy explains that U.S. government works are generally public domain and that embedded third-party material can retain copyright restrictions. This Actor returns structured observations and official source links; a publication's availability does not imply that every embedded work has the same reuse status.

The output contains corporate party names and recorded law-firm evidence, not personal attorney names, phone numbers, or email addresses. The Actor performs no outreach. Consult the cited source when a relationship or procedural role matters to your research.

Version 0.2 input and data migration

Version 0.2 uses GovInfo case packages and published opinions. docket_id is now a string package ID. Use opinion_date_from and opinion_date_to for opinion dates, and maxSourceRequests for the source-request limit. Legacy filing_date_from, filing_date_to, and maxApiRequests keys are rejected rather than reinterpreted. Source-token input fields are no longer accepted.

Update existing integrations to treat filing_date as unavailable, handle nullable Nature of Suit fields, and read the new opinion-coverage and firm-evidence fields. Existing source-specific docket identifiers and historical representation links are not interchangeable with the new package and opinion identifiers.

Resume time accounting

The maxRunSeconds deadline is fixed for the same run. Time spent interrupted or waiting for a restart counts toward this wall-clock limit; resuming does not reset it. Request counts and provider retry deadlines are also preserved. The recorded case_name retains the public source case title; corporate-party and firm fields exclude personal attorney names and contact details.

The processing deadline bounds source work and storage operations. No new Dataset append starts after it expires. A final checkpoint and operational-summary write each have a separate maximum of 10 seconds for shutdown; use Apify's run timeout for the platform wall-clock ceiling. A timed-out append remains pending and is checked against the Dataset on resume before another append is attempted.

For an append whose server outcome is unknown, resuming first checks the Dataset. If the record is present, it is reconciled without another append. If it is still absent, the run stops with delivery_unconfirmed and retains the pending record instead of risking a second paid row. Dataset writes use a single HTTP attempt; source downloads retain their separate controlled retries.