Launch Week 3: Five days of launches
Confident AI
Docs
Pricing
Products
LLM Evaluation
LLM Evaluation

Benchmark LLM systems with research-backed metrics.

LLM Observability
LLM Observability

Trace, monitor, and alert on production LLM systems.

AI Red Teaming
AI Red Teaming

Stress-test LLM apps against adversarial attacks.

AI Governance
AI Governance

Enforce AI standards and controls across teams.

Book a DemoSign Up
Book a DemoSign Up

Goal Accuracy

Goal Accuracy is a multi-turn metric to determine if your agent reached the user's goal

Overview

The goal accuracy metric is a multi-turn metric that uses LLM-as-a-judge to evaluate whether your LLM agent planned and executed its way to the goal the user was after across the conversation.

Required Parameters

These are the parameters you must supply in your test case to run evaluations for goal accuracy metric:

turnslist of TurnRequired

A list of Turns as exchanges between user and assistant.

Parameters of Turn:

roleuser | assistantRequired

The role of the person speaking, it's either user or assistant

contentstringRequired

The content provided by the role for the turn

How Is It Calculated?

The goal accuracy metric first extracts the goal of each user and assistant interaction along with the steps your agent took, then uses an LLM to score each goal-steps pair on whether the goal was reached and on the quality of the plan behind it.


Goal Accuracy=Goal Evaluation Score+Plan Evaluation Score2\text{Goal Accuracy} = \frac{\text{Goal Evaluation Score} + \text{Plan Evaluation Score}}{2}

The final score is the average of the goal evaluation score and the plan evaluation score across the conversation.

Usage

To run the goal accuracy metric on Confident AI, add it to a multi-turn metric collection. This will allow you to use goal accuracy metric for:

  • Multi-turn E2E testing
  • Online and offline evals for traces and spans

Last updated on

Built byConfident AI