AmpleData Enrichment — Apify Actor

Enrich a CSV through the AmpleData API: upload a list, define the columns you want filled in plain English, and get back cited, confidence-scored cells.

The actor runs one enrichment job and is cancellable — aborting the Apify run cancels the underlying AmpleData job via POST /jobs/{jobID}/cancel.

API key

Two ways to supply the AmpleData key — the actor uses whichever is present:

1. Bring your own key — set the apiToken input. Enrichment is billed per cell against your AmpleData balance. Generate a key at https://ampledata.io account settings.
2. Keyless — leave apiToken empty. The actor falls back to the AMPLEDATA_KEY environment variable configured on the actor, so users run without a key (these runs are billed by the actor owner, typically via Apify monetization).

apiToken always wins when both are set. The owner key lives in an Apify secret environment variable (Console → Actor → Settings → Environment variables → AMPLEDATA_KEY, marked Secret) — never hardcoded in source.

Input

Field	Required	Description
apiToken	yes	AmpleData API key (sk_live_...). Generate from account settings.
columns	yes	Columns to enrich. Each: { name, type, description }. type ∈ string|number|boolean|date.
csv	one of	Inline CSV (first row = header).
csvUrl	one of	URL of a CSV to download. Takes precedence over csv.
keyColumns	no	Identifying columns (e.g. ["company"]). Empty → AmpleData picks.
keyColumnDescription	no	Plain-English description of the key column.
rowLimit	no	Max rows to process.
baseUrl	no	API base URL. Default https://ampledata.io/api/v1.
pollIntervalSecs	no	Progress poll interval. Default 5.
insecureTLS	no	Skip TLS verification. Only for a self-hosted endpoint with a self-signed cert. Insecure.

Example

{
  "apiToken": "sk_live_...",
  "csv": "company\nstripe.com\nfigma.com",
  "keyColumns": ["company"],
  "columns": [
    { "name": "industry", "type": "string", "description": "Primary industry" }
  ]
}

Output

Each enriched row is pushed to the default dataset:

{
  "key": "stripe.com",
  "extracted_data": { "industry": "Fintech" },
  "confidence": { "industry": { "score": 0.9, "reason": "..." } },
  "sources": ["https://..."]
}

A summary is written to the key-value store under OUTPUT: { jobId, status, totalRows, resultCount }.

How it works

1. POST /enrichment-signed-url → signed upload URL + sourceId
2. PUT the CSV bytes to the signed URL
3. POST /sources/{sourceId}/enrich → jobId
4. Poll GET /jobs/{jobId}/progress until COMPLETED / CANCELLED
5. GET /jobs/{jobId}/results → dataset

Develop

npm install
npm run build
apify run    # uses storage/key_value_stores/default/INPUT.json

Deploy

apify login
apify push