Canonical
GET /api/enrichment
Pipeline view of domain enrichment — who has a verified domain, who's ambiguous, and why.
Data source: Orakel's domain-enrichment pipeline. Maps Norwegian companies to web domains, LinkedIn, Facebook, technologies, and ad pixels. This endpoint exposes the pipeline state — enriched fields themselves are embedded on company records.
Request
GET /api/enrichment
| Parameter | Type | Required | Description |
|---|---|---|---|
status |
"confirmed" | "ambiguous" | "rejected" | "pending" |
no | Default: all statuses except pending. |
minConfidence |
integer | no | Default 0. 0–100 scale. |
limit |
integer | no | 1–200. Default 50. |
offset |
integer | no | Default 0. Offset pagination, not cursor. |
Response
{
"summary": { "confirmed": 120000, "ambiguous": 8000, "rejected": 4500 },
"total": 132500,
"limit": 50,
"offset": 0,
"data": [
{
"orgNumber": "912345678",
"companyName": "Example AS",
"status": "confirmed",
"confidence": 92,
"primaryDomain": "example.no",
"linkedinHandle": "example-as",
"facebookHandle": null
}
]
}Each row also includes confirmedDomains[], candidates[] (the full scoring ledger), stage, completedAt, brregWebsite (self-reported), and basic firmographics (orgForm, city, employees). Pipeline mechanics and confidence thresholds are documented in the enrichment source page.
Examples
curl
curl -H "Authorization: Bearer $ORAKEL_KEY" \
"https://orakel.cloud/api/enrichment?status=ambiguous&minConfidence=40&limit=100"JavaScript
const res = await fetch(
"https://orakel.cloud/api/enrichment?status=ambiguous&limit=100",
{ headers: { Authorization: `Bearer ${process.env.ORAKEL_KEY}` } },
);
const { summary, data } = await res.json();Error modes
| Status | Meaning | Action |
|---|---|---|
| 401 | Missing or invalid API key | Check the Authorization header |
| 429 | Rate or quota exceeded | Back off |
Rate limits
Per-key RPM per tier.
Typical use cases
- QA: audit
ambiguousrows and their candidate lists before promoting toconfirmed. - Coverage reporting: read
summaryto track enrichment completion over time. - Targeted re-runs: identify low-confidence rows for manual review.