jeff.support.triagev1.4.0

Support ticket triage

Routes a support ticket or email to a team and assesses urgency, sentiment and the need for human review.

Example

Context

company: An online shop selling outdoor clothing and camping gear across Europe.
channel: email

Input

Subject: Wrong item in my parcel Hi, my order 88213 arrived today but the parcel had a large Alpine shell jacket instead of the medium I ordered. Please exchange it and send the medium before Saturday. Thanks, Marta

Labels (the options)

orders: Orders, deliveries, returns and exchanges
payments: Payments, refunds and invoices
product: Product questions, sizing advice and stock
account: Customer accounts, passwords and newsletters

Answer (JSON)

value orders, confidence 1

Top options

orders 1
product 0
payments 0

A wrong item to exchange is a matter for the orders team.

Open in playground

import os import jeff with jeff.Jeff.remote("https://api.jeffhub.ai", token=os.environ["JEFF_API_KEY"], timeout_s=30) as client:     decide = client.function("jeff.support.triage.route", context={        "company": "An online shop selling outdoor clothing and camping gear across Europe.",        "channel": "email",    }, labels={        "orders": "Orders, deliveries, returns and exchanges",        "payments": "Payments, refunds and invoices",        "product": "Product questions, sizing advice and stock",        "account": "Customer accounts, passwords and newsletters",    })     decision = decide("Subject: Wrong item in my parcel\nHi, my order 88213 arrived today but the parcel had a large Alpine shell jacket instead of the medium I ordered. Please exchange it and send the medium before Saturday. Thanks, Marta")     print(decision.value, decision.confidence)

More examples

Charged twice for one orderpayments · 1

Context

company: An online shop selling outdoor clothing and camping gear across Europe.
channel: email

Input

Subject: Double charge Hello, my card was charged twice for order 88340, 129.00 each time, but I only placed one order. Please refund the extra payment. Regards, Henrik

Labels (the options)

orders: Orders, deliveries, returns and exchanges
payments: Payments, refunds and invoices
product: Product questions, sizing advice and stock
account: Customer accounts, passwords and newsletters

Answer (JSON)

value payments, confidence 1

Top options

payments 1
orders 0
product 0

A card charged twice for one order goes to the payments team.

Open in playground

import os import jeff with jeff.Jeff.remote("https://api.jeffhub.ai", token=os.environ["JEFF_API_KEY"], timeout_s=30) as client:     decide = client.function("jeff.support.triage.route", context={        "company": "An online shop selling outdoor clothing and camping gear across Europe.",        "channel": "email",    }, labels={        "orders": "Orders, deliveries, returns and exchanges",        "payments": "Payments, refunds and invoices",        "product": "Product questions, sizing advice and stock",        "account": "Customer accounts, passwords and newsletters",    })     decision = decide("Subject: Double charge\nHello, my card was charged twice for order 88340, 129.00 each time, but I only placed one order. Please refund the extra payment. Regards, Henrik")     print(decision.value, decision.confidence)

Which sleeping bag for winter?product · 1

Context

company: An online shop selling outdoor clothing and camping gear across Europe.
channel: web form

Input

Is the Fjellbu 600 sleeping bag warm enough for camping in the mountains in February, around minus 10? Or should I take the 900?

Labels (the options)

orders: Orders, deliveries, returns and exchanges
payments: Payments, refunds and invoices
product: Product questions, sizing advice and stock
account: Customer accounts, passwords and newsletters

Answer (JSON)

value product, confidence 1

Top options

product 1
orders 0
account 0

A question about which sleeping bag is warm enough is for the product team.

Open in playground

import os import jeff with jeff.Jeff.remote("https://api.jeffhub.ai", token=os.environ["JEFF_API_KEY"], timeout_s=30) as client:     decide = client.function("jeff.support.triage.route", context={        "company": "An online shop selling outdoor clothing and camping gear across Europe.",        "channel": "web form",    }, labels={        "orders": "Orders, deliveries, returns and exchanges",        "payments": "Payments, refunds and invoices",        "product": "Product questions, sizing advice and stock",        "account": "Customer accounts, passwords and newsletters",    })     decision = decide("Is the Fjellbu 600 sleeping bag warm enough for camping in the mountains in February, around minus 10? Or should I take the 900?")     print(decision.value, decision.confidence)

Stove that caught firetrue · 1

Context

company: An online shop selling outdoor clothing and camping gear across Europe.
channel: email

Input

Subject: Your stove set my tent on fire The gas stove I bought from you (order 87112) flared up on its first use and burned a hole in my tent. My hand was burned too. I want to know what you are going to do about this.

Answer (JSON)

value true, confidence 1

Top options

true 1
false 0

A stove that set a tent on fire and burned the customer is a safety complaint, so it needs a person and the answer is true.

Open in playground

import os import jeff with jeff.Jeff.remote("https://api.jeffhub.ai", token=os.environ["JEFF_API_KEY"], timeout_s=30) as client:     decide = client.function("jeff.support.triage.needs_human", context={        "company": "An online shop selling outdoor clothing and camping gear across Europe.",        "channel": "email",    })     decision = decide("Subject: Your stove set my tent on fire\nThe gas stove I bought from you (order 87112) flared up on its first use and burned a hole in my tent. My hand was burned too. I want to know what you are going to do about this.")     print(decision.value, decision.confidence)

Password reset questionfalse · 1

Context

company: An online shop selling outdoor clothing and camping gear across Europe.
channel: chat

Input

hi how do I reset my password? the link on the login page is where?

Answer (JSON)

value false, confidence 1

Top options

false 1
true 0

Where to find the password reset link can be answered automatically, so the answer is false.

Open in playground

import os import jeff with jeff.Jeff.remote("https://api.jeffhub.ai", token=os.environ["JEFF_API_KEY"], timeout_s=30) as client:     decide = client.function("jeff.support.triage.needs_human", context={        "company": "An online shop selling outdoor clothing and camping gear across Europe.",        "channel": "chat",    })     decision = decide("hi how do I reset my password? the link on the login page is where?")     print(decision.value, decision.confidence)

Use when

Each ticket, email or chat message needs a team assignment, urgency rating and human-review decision.
You want to use your own team list, with a short description of each team.

Avoid when

You need to write a reply; this adapter returns decisions without generating text.
The decision depends on facts outside the message, such as order history; include them in the request or handle them in code.

Context and input

Field	Kind	Description
message	input	The ticket or email as received, including subject line, quoted thread and signature.
company	context	One sentence about the organisation the message is sent to.
channel	context	Where the message came from, for example email, web form, chat, app review, social media reply or phone transcript.

Functions

Key	Interface	Instructions	Labels
route	route	Which team should handle this message? If it covers several issues, choose the team for the most important one.	from the caller
urgency	rate	How urgently does this need a response?	No action needed, Can wait a few days, Handle within a day, Handle within hours, Handle immediately
sentiment	rate	How does the customer feel?	Very negative, angry or distressed, Negative, Neutral, Positive, Very positive
needs_human	requires_escalation	Does this need a person to act on it now, rather than an automatic reply? Answer yes for complaints that could escalate, legal or safety issues, or requests an automatic system cannot resolve.	false, true

How it was measured

Test accuracy: the share of test rows whose top answer was right.

91.2% on the in-house test (7,256 rows). in-house: same generator as the training data

When confident

From the calibration set: at or above the threshold confidence, how often the answer is right, and how often the adapter is that confident. Below it, escalate.

Function	Right when confident	Confident on	Threshold	Calibration rows
route	95.1%	100.0%	0.43	367
urgency	95.0%	59.9%	0.79	367
sentiment	95.0%	88.0%	0.69	367
needs_human	98.9%	100.0%	0.54	367

Base: Jeff v1.4. Adapter trained: 25 October 2026.

Base

jeff-base v1.4

Call it

jeff.support.triage.route, with placeholders for its input and context.

import os import jeff with jeff.Jeff.remote("https://api.jeffhub.ai", token=os.environ["JEFF_API_KEY"], timeout_s=30) as client:     decide = client.function("jeff.support.triage.route", context={        "company": "<company>",        "channel": "<channel>",    })     decision = decide("<message>")     print(decision.value, decision.confidence)

Download

Weights: mstrasser/jeff-adapter-triage (v1.4)

Licence: Apache-2.0