Create a Classification Function

This is the first step in building a classification function with your own labels and data. The next four pages walk through defining labels, invoking predictions, reviewing them, and where to go next.

What you’ll build

A classification function takes an input (text, image, or structured data) and returns one of your labels with a confidence score. By the end of this five-page guide you’ll have:

• A live endpoint
• Your own labels defined
• Predictions flowing in
• A review workflow that automatically trains a private model on your data
• Code calling the function from your application

Total time: ~15 minutes (plus however long it takes to gather your first examples).

1. Open the console and create a function

Go to the Nyckel console and sign in. Click New function.

2. Pick an input type

Choose based on what your application will actually send to the function:

• Text — written content: messages, documents, support tickets, product descriptions, reviews.
• Image — photos, screenshots, scanned documents, product images.
• Tabular — structured rows of numeric or categorical data (e.g. sensor readings, form submissions).

The input type is set at creation time and can’t be changed later, so pick the one that matches your real production input.

3. Pick the function type

Select Classification. (This guide is the classification path; for Box Detect or Search, see Function Types.)

4. Name the function

Use a short, descriptive name that reflects what the function classifies. Lowercase and hyphens work well for API use. A good name describes the input and the decision — for example:

• support-ticket-type
• product-image-category
• review-sentiment

Click Create. Your function is live immediately. You now have a real endpoint with a unique ID, even though it doesn’t know your labels yet.

What you just did

Next

Define Labels — set up the decision space your function will use.