French Companies Enricher (SIREN lookup for KYC & B2B prospecting)

Turn a list of French company names, SIRENs or SIRETs into a clean, structured dataset of legal registry records — identity, head-office address, activity, headcount bracket, officers and public compliance flags.

Data source: recherche-entreprises.api.gouv.fr, the official free open-data search API over the French national business register (SIRENE / INSEE, RNE, INPI and partner datasets). No API key, no proxy, no login.

Typical buyers:

Buyer	What they do with it
B2B sales / lead-gen teams	Append SIREN, legal name, NAF code, headcount bracket and officers to a prospect list; filter out closed companies.
KYC / KYB & compliance teams	Verify that a counterparty exists, is active (etat_administratif), who its registered officers are, and whether it is an association, ESS or mission-driven company.
Due diligence / M&A analysts	Check legal form, creation and closing dates, sites, published revenue for large companies.
Data teams	Resolve messy company names to canonical SIREN identifiers before joining internal data.

---

Input

{
  "queries": ["552100554", "542051180", "boulangerie paris"],
  "maxResultsPerQuery": 25,
  "includeDirigeants": true,
  "includeEtablissements": false
}

Field	Type	Default	Description
queries	array of strings	— (required)	Company names, 9-digit SIRENs, or SIRETs. The registry needs at least 3 characters per term; a term shorter than that is skipped with a warning instead of failing the run.
maxResultsPerQuery	integer 1–500	25	Cap of companies kept per query, and therefore your cost cap: one company = one billed event. Name searches can match hundreds of companies (e.g. boulangerie paris → 988 matches), so this is what keeps a run cheap.
includeDirigeants	boolean	true	Adds the dirigeants array (people and corporate officers with role, birth year/month, SIREN for legal entities).
includeEtablissements	boolean	false	Adds the etablissements array of matching secondary sites (SIRET, address, activity, dates, coordinates).

Each company is emitted once per run: if two queries match the same SIREN, the duplicate is dropped before billing.

---

Output

One dataset item per enriched company. Full example (real record, trimmed):

{
  "siren": "552100554",
  "siren_conserve": null,
  "nom_complet": "PEUGEOT SA",
  "nom_raison_sociale": "PEUGEOT SA",
  "sigle": null,
  "etat_administratif": "C",
  "date_creation": "1955-01-01",
  "date_fermeture": "2021-01-16",
  "nature_juridique_code": "5699",
  "categorie_entreprise": null,
  "tranche_effectif_salarie_code": "NN",
  "annee_tranche_effectif_salarie": null,
  "activite_principale_naf": "70.10Z",
  "activite_principale_naf25": null,
  "section_activite_principale": "M",
  "nombre_etablissements": 5,
  "nombre_etablissements_ouverts": 0,
  "tva": null,
  "siege": {
    "siret": "55210055400054",
    "adresse": "RTE DE GIZY 78140 VELIZY-VILLACOUBLAY",
    "complement_adresse": null,
    "code_postal": "78140",
    "commune_code_insee": "78640",
    "commune": "VELIZY-VILLACOUBLAY",
    "departement": "78",
    "region": "11",
    "epci": "247800584",
    "pays_etranger": null,
    "latitude": "48.774126458",
    "longitude": "2.2219642076",
    "activite_principale": "70.10Z",
    "etat_administratif": "F",
    "date_creation": "2020-04-27",
    "date_fermeture": "2021-01-16",
    "tranche_effectif_salarie_code": "NN",
    "nom_commercial": null,
    "enseignes": null
  },
  "complements": {
    "est_entrepreneur_individuel": false,
    "est_association": false,
    "est_service_public": null,
    "est_administration": false,
    "egapro_renseignee": true,
    "est_ess": false,
    "est_societe_mission": false,
    "est_organisme_formation": false,
    "est_qualiopi": false,
    "convention_collective_renseignee": false,
    "liste_idcc": null,
    "est_avocat": false,
    "est_siae": false,
    "type_siae": null,
    "est_patrimoine_vivant": false,
    "identifiant_association": null
  },
  "finances": { "2019": { "ca": 74731000000, "resultat_net": 3584000000 } },
  "query": "552100554",
  "source_url": "https://annuaire-entreprises.data.gouv.fr/entreprise/552100554",
  "source_api": "https://recherche-entreprises.api.gouv.fr/search?q=552100554",
  "retrieved_at": "2026-10-26T19:12:44+00:00",
  "dirigeants": [
    {
      "type_dirigeant": "personne physique",
      "nom": "ANTUNES TAVARES DIAS",
      "prenoms": "CARLOS",
      "annee_de_naissance": "1958",
      "date_de_naissance": "1958-08",
      "qualite": "Président du directoire",
      "nationalite": null,
      "siren": null,
      "denomination": null
    }
  ],
  "confidence": 0.95
}

Field reference

Field	Type	Description
siren	string	9-digit national company identifier. Stable join key.
nom_complet	string	Legal name as published (falls back to nom_raison_sociale, then the SIREN itself if the registry record is empty).
nom_raison_sociale, sigle	string | null	Registered name and acronym.
etat_administratif	"A" | "C" | null	Company status: A = active, C = ceased. Filter on this before prospecting.
date_creation, date_fermeture	ISO date | null	Creation and, when applicable, closing date.
nature_juridique_code	string | null	INSEE legal-form code (e.g. 5699 = société européenne).
categorie_entreprise	PME/ETI/GE | null	Size category, published only for part of the register.
tranche_effectif_salarie_code	string | null	Headcount bracket code (INSEE scale, e.g. 02, 11, 51; NN = not available).
annee_tranche_effectif_salarie	string | null	Year the headcount bracket refers to.
activite_principale_naf	string | null	NAF activity code of the company.
activite_principale_naf25, section_activite_principale	string | null	NAF 2025 code when published, and the NAF section letter.
nombre_etablissements, nombre_etablissements_ouverts	integer	Total and currently open sites.
tva	string | null	Intra-EU VAT number when published.
siege	object	Head office: siret, adresse, complement_adresse, code_postal, commune (city name), commune_code_insee, departement, region, epci, pays_etranger, latitude, longitude, activite_principale, etat_administratif, date_creation, date_fermeture, tranche_effectif_salarie_code, nom_commercial, enseignes.
complements	object	Public flags: est_entrepreneur_individuel, est_association, est_service_public, est_administration, egapro_renseignee, est_ess, est_societe_mission, est_organisme_formation, est_qualiopi, convention_collective_renseignee, liste_idcc, est_avocat, est_siae, type_siae, est_patrimoine_vivant, identifiant_association.
finances	object | null	Published accounts by year: { "2019": { "ca": …, "resultat_net": … } }. Only for companies that file public accounts.
dirigeants	array	Present when includeDirigeants is true. Each entry: type_dirigeant (personne physique / personne morale), nom, prenoms, annee_de_naissance, date_de_naissance (YYYY-MM), qualite, nationalite, siren, denomination.
etablissements	array	Present when includeEtablissements is true. Each entry: siret, adresse, code_postal, commune, commune_code_insee, departement, region, epci, activite_principale, etat_administratif, est_siege, date_creation, date_fermeture, tranche_effectif_salarie, annee_tranche_effectif_salarie, latitude, longitude, nom_commercial, enseignes.
query	string	The input query that produced this record.
source_url	string	Human-verifiable page on the official annuaire-entreprises.data.gouv.fr.
source_api	string	Exact API request used, for reproducibility.
retrieved_at	ISO 8601	Fetch timestamp (UTC).
confidence	number 0–1	Data-completeness score, not a probability that the company exists. It is the weighted share of key registry fields actually populated (name, SIRET, postal code, address, NAF, creation date, status, city, legal form, headcount bracket), ignoring the registry's NN "unknown" sentinel. A sparse record (unknown SIREN) scores low; a fully documented one scores 1.0.

---

Pricing

Suggested pay-per-event pricing: one event company-enriched per dataset item.

Volume	Price per event	Price per 1,000 results
Suggested	$0.002	$2.00
Low-volume / high-touch	$0.005	$5.00
High-volume data teams	$0.001	$1.00

The underlying API is free and open data, so the price covers the enrichment, normalisation, pagination, retries, deduplication and the stable schema (codes normalised, num→string, nulls kept explicit, commune split into code + name). Comparable Apify actors in the company-data niche charge $1–$5 per 1,000 records. Runs with zero results are never billed.

---

Known limits (verified against the live API)

• No human-readable labels for codes. The API returns nature_juridique (5699), tranche_effectif_salarie (NN, 02…), departement (78) and region (11) as codes only — no libelle_* for these, so this Actor does not invent one. Bring your own lookup table if you need "SARL", "50–99 employees" or "Île-de-France". City names are available (siege.commune), because the API's commune field is the INSEE code and its libelle_commune is the name — this Actor exposes both.
• per_page is capped at 25 by the API (per_page=100 → HTTP 400), so results above 25 require extra paged requests. Deep pagination is capped at 40 pages per query.
• Queries must be at least 3 characters. Shorter terms return HTTP 400; the Actor logs a warning, records error: "invalid_query" in the run summary and continues.
• Empty records exist. An unknown SIREN can return a shell record with a SIREN and dates but no name or address. Those items keep nom_complet = the SIREN, carry no fabricated values, and score a low confidence (< 0.5) so you can filter them out.
• Officers are partial. Only published dirigeants appear, and natural persons may opt out of public diffusion — nom can be null. Roles are in French, as published (Président du directoire, Gérant, Commissaire aux comptes…).
• etablissements only lists matching sites, not the full site list of the company. For the full list, use the company's siren with the same API's q= search.
• finances is sparse — only companies filing public accounts have it, and only for the years the registry exposes.
• No rate-limit headers are published. This Actor self-limits to 10 requests/second and retries twice (honouring Retry-After when present). HTTP 429 was observed during testing under burst load and was recovered transparently.
• Public-sector / foreign entities are only present insofar as the register carries them; est_service_public is frequently null rather than false.

---

Run locally

python3 -m venv .venv && .venv/bin/pip install -r requirements.txt
INPUT_JSON='{"queries":["552100554","danone"],"maxResultsPerQuery":5}' \
  .venv/bin/python -m src.main
# results land in ./storage/datasets/default/
.venv/bin/python -m pytest tests -q

INPUT_JSON is a local-development convenience only; on the Apify platform the Actor reads its input from the standard INPUT key-value record.

Layout

.actor/actor.json          Actor definition (input/output/dataset schema refs)
.actor/input_schema.json    Input JSON Schema (English, buyer-facing)
.actor/output_schema.json   Dataset + run-summary links
.actor/dataset_schema.json  Table views (overview, KYC)
actor -> .actor            Convenience symlink to the same files
src/main.py                 The Actor (stdlib HTTP + apify SDK, no third-party client)
Dockerfile                  apify/actor-python:3.12
tests/test_fields.py        Field-contract + exit-code tests against the live API
proof/                      Real run output and measurements

Publishing (steps left for the account owner)

1. npm i -g apify-cli && apify login (a personal API token is required — not available on the machine this Actor was built on).
2. apify push from this directory.
3. In the Console → Monetization, enable pay-per-event with event company-enriched at $0.002, matching the runs' charged event name.
4. Publish to the Store with the suggested categories Lead generation, Business data, KYC / compliance.