Merino crew neck sweater, navy, sizes S to XL
Apparel › Knitwear
Where classification usually starts
Most teams sort text with a frontier model and a prompt that lists every category with notes on the hard cases. As the list grows, the prompt gets longer, answers drift, and the cost per item stays high. A specialist knows your categories and sorts every item by the same rules.
What people sort with it
- Catalogs: product listings into your category tree
- Inbound: messages by intent, such as sales, support or partnership
- Leads: fit for your ideal customer, with a short reason
- Moderation: posts against your own policy, with the rule each one breaks
- Feedback: reviews and survey answers by topic and sentiment
How it works
- Describe the job. List your categories, or paste the prompt you use today if you have one, and add a few real items. Facet writes and keeps up the instructions from there.
- Agree on what good looks like. Facet writes plain statements of a good answer, like “anything knitted is Knitwear, whatever the garment”, each with examples. Edit any that are off.
- Create, then correct. Facet builds the specialist on its own, usually in about half an hour, and checks it against every statement before it goes live. When an item lands in the wrong category, say what should change. Facet quotes the fix before it spends anything.
Three ways to use it
- In the app: sign in, describe the categories, and choose Create.
- From your agent: connect Claude, Cursor, ChatGPT, Codex or VS Code to Facet, then paste:
Use Facet to build a specialist that sorts our product listings into this category tree:
... Here are twenty real listings: ... Show me the draft and the price before you approve
anything.
- In your code: keep your OpenAI SDK and change two lines, the address and the model name.
import os
from openai import OpenAI
client = OpenAI(base_url=os.environ["FACET_BASE_URL"], api_key=os.environ["FACET_API_KEY"])
reply = client.chat.completions.create(model="product-categories", messages=[{"role": "user", "content": listing}])
Why a specialist for classification
- Steady answers. A new version goes live only after it passes every statement you agreed on, and any update rolls back in one step.
- Fast and low cost. A Facet specialist does one job, so it answers faster than a frontier model, and every build shows its price before it starts.
- Better with every correction. Each item you flag becomes a check that every later version must pass.
- Yours. Export it any time, free, and run it wherever you like.
Questions
Can a specialist give more than one label?
Yes. Describe it that way, such as "tag each review with every topic it mentions", and Facet returns a list of labels in the same shape on every call.
What if my categories change?
Ask for the change in plain words, such as "add a Bags category for totes and backpacks", with a few examples. Facet quotes the change, builds it, and checks the new version against every rule before it goes live.
Is a specialist worth it for a few hundred items a month?
Usually not. At low volume a frontier model with a good prompt is simpler. A specialist pays off when the same job runs thousands of times, or when speed and steady answers matter.