# Industry & SIC Classifier

A domain or name → its industry, SIC codes and what it does.

Point it at a business and get back the official SIC codes, a plain-English industry label, a one-line description of what they do, and a confidence rating.

Industry classification sounds solved until you try it: businesses mislabel themselves, registries lag reality, and a bare domain tells you nothing. This endpoint reads the register AND the business's own website, then returns both the official SIC codes and a plain-English label with a one-line description — plus an honest confidence rating instead of a silent guess. Consistent categorisation, at machine speed, from the two sources that actually matter.

- Price: 2 credits
- Page: https://meetkatalis.com/apis/industry-classifier
- OpenAPI: https://meetkatalis.com/openapi.yaml

## Endpoints

- `GET /api/v1/industry-classifier?q=dishoom.com` — Industry + SIC + confidence

## Call it

```bash
curl -H "x-api-key: $KATALIS_API_KEY" \
  "https://meetkatalis.com/api/v1/industry-classifier?q=greggs.com"
```

```javascript
const res = await fetch("https://meetkatalis.com/api/v1/industry-classifier?q=greggs.com", {
  headers: { "x-api-key": process.env.KATALIS_API_KEY },
});
if (!res.ok) throw new Error(`Katalis ${res.status}: ${await res.text()}`);
console.log(await res.json());
```

```python
import os, requests

r = requests.get(
    "https://meetkatalis.com/api/v1/industry-classifier?q=greggs.com",
    headers={"x-api-key": os.environ["KATALIS_API_KEY"]},
    timeout=60,
)
r.raise_for_status()
print(r.json())
```

## Example response (a real reply)

```json
{
  "input": "dishoom.com",
  "industry": "Restaurants",
  "sicCodes": ["56101"],
  "whatTheyDo": "Bombay-style café/restaurant group.",
  "confidence": "high"
}
```

## What you'd build with it

- **Segmentation that's real.** Tag every account with official SIC codes + a human label — your analytics and routing stop guessing industries.
- **Risk & underwriting buckets.** Fintech and insurance flows that need 'what does this business actually do' answered consistently, at machine speed.
- **Marketplace categorisation.** Sellers self-categorise wrongly; this classifies them from their own website and the register instead.

## Why this one

- **Plain-English + official codes.** Both the human label and the SIC codes, from one call — not just a raw registry code.
- **Grounded, with confidence.** Backed by a verified company match and an honest confidence rating, not a guess.

## Conventions

- Auth: `x-api-key: kat_live_…` on every request (free key at https://meetkatalis.com/developers: a verified email and mobile, 100 credits a month, no card).
- One credit balance across every API; each endpoint's price is on its page and in `x-credits` in https://meetkatalis.com/openapi.yaml.
- Keys are camelCase everywhere. A registered company is `company: { name, number, status, statusDetail?, filingsOverdue? }` on the register endpoints; the enrichment family (enrich, score, assess, brand-kit, ai-visibility, tender-match) keeps `business` / `profile` with `companyNumber`.
- Every error has `code` and `requestId`; every reply an `x-request-id` header; a charged reply `x-credits-remaining`; a reply that gave credits back `x-credits-refunded`. Missing = 404 with `query`.
- Burst ceiling 60 requests a minute per key (429, `Retry-After`, no credits used). A failure on our side (5xx) gives the call's credits back.
- Company data contains public sector information licensed under the Open Government Licence v3.0.

Provided "as is"; results come from public sources and, in places, AI, and may be inaccurate or out of date; not legal, professional or credit advice. Terms: https://meetkatalis.com/apis/terms