AI Evaluation

Classify, score and grade any record with a decision model.

Decision models built for classification. Ask a typed question about any record and get a calibrated probability back, inside your pipelines and next to your data.

Start a pilot Join the waitlist

Typed questions, typed answers

Ask typed questions (a choice, a score, a probability) about any row, such as a lead, a session or a support ticket.

Answers as probabilities

Every answer comes back as a probability you can calibrate against real outcomes in your own warehouse.

Edit the standard in plain text

Criteria are written in plain language. Change the text and the next run applies the new definition without retraining.

Inside your pipelines

Runs inside managed pipelines, with concurrency, retries and every answer stored next to the data it judged.

In code

pipeline/ingestion/evaluation/lead_grades.py

client = s.evaluation_client("evaluation")
result = client.evaluate(
    state={"job_title": "Head of L&D", "company": "Acme"},
    questions={
        "grade": {
            "type": "choice",
            "instructions": "Which grade does this demo request match?",
            "criteria": {"A": "...", "B": "...", "C": "...", "D": "..."},
        },
    },
)
grade = result["answers"]["grade"]

Pipeline SDK reference →

Use it, install it, or build on it

Every block on the platform comes three ways. Most teams use all three.

Use it

Declare the evaluation connection; there is nothing to configure.

Install it

Install lead scoring or traffic quality, which come with evaluation wired in.

Build on it

Your agent writes the questions and criteria for any record your business needs to judge.

Replaces Per-client ML training pipelines and Brittle scoring rules.

Install from the marketplace

Jev evaluation

A decision model for classifying and scoring records.

AI model · TypeSafe AI

Lead Scoring

Grade every demo request A–D against standards you can read and edit.

App

Traffic Guard

Score paid traffic for bots and invalid clicks.

App

Churn Risk

Health scores and churn probability, with the reasons.

App

Support Analytics

Why customers contact you, and how fast you answer.

App

Browse the marketplace →

Solutions built with it

Lead scoring

Grade leads A to D against standards you can read

Traffic quality

Score paid traffic for bots and invalid clicks

Enterprise pilot

Start a pilot.

Want ai evaluation set up for you? Our forward-deployed engineers build it for you in your workspace, as code you own.

One solution, on your data, in your own workspace
First version live in two weeks
Parity proven before anything is switched off
Fixed scope and price, credited to your contract if you continue

Not ready for a pilot? Send the form anyway with your question. An engineer will reply, not a sales rep.

How the pilot works

Your company’s context, ready for every agent.

$ segmentstream workspace create acme --region eu

▸ workspace ready · data, objects, metrics and access rules in one place

▸ connect Claude, ChatGPT or Cursor with one connector

▸ tell your agent: “add our billing API as a source”

Start a pilot

Our engineers build the first version on your data in two weeks, at a fixed price.

Or join the waitlist for self-serve:

Pilots are open now. Self-serve opens to the waitlist this fall.