d1-ultra Choice
How to run d1-ultra Classify a customer inquiry
Description
A Choice from Liquid’s usage examples: closed labels, a distribution, and a confidence value.
Copy
import os
from typesafe_sdk import Choice, Noul, Score, TypeSafeClient
client = TypeSafeClient(
api_key=os.environ["LIQUID_API_KEY"],
base_url="https://api.liquid.ai",
)
result = client.system_one(
model="d1-ultra",
state="Hi, I was charged twice for my subscription last month. The amounts are $29.99 on March 3rd and again on March 5th. Can you refund the duplicate charge?",
questions={
"intent": Choice(
instructions="What is the primary topic of this customer inquiry?",
criteria={
"billing": "Questions about charges, invoices, refunds, or payment methods",
"technical": "Bug reports, feature requests, or how-to questions about the product",
"shipping": "Order status, delivery tracking, or shipping address changes",
"returns": "Return requests, exchanges, or product condition issues",
"account": "Login issues, profile updates, or subscription management"
},
)
},
)
Worked example
Illustrative for this sample state, not a live run
Sample state
Hi, I was charged twice for my subscription last month. The amounts are $29.99 on March 3rd and again on March 5th. Can you refund the duplicate charge?
Choice
intentWhat is the primary topic of this customer inquiry?
- billing · selected 100%
Questions about charges, invoices, refunds, or payment methods
- technical 0%
Bug reports, feature requests, or how-to questions about the product
- shipping 0%
Order status, delivery tracking, or shipping address changes
- returns 0%
Return requests, exchanges, or product condition issues
- account 0%
Login issues, profile updates, or subscription management
Confidence 100%
What you might do next
Map `intent` onto a queue (`billing-team`, `engineering-support`, `logistics`, `returns-desk`, `account-management`) and log the confidence.
Tags
choicesupport