Analyzing ticket conversation
Evaluating against rubric dimensions
Gathering evidence for each score

The AI is carefully reviewing the entire conversation to provide accurate, evidence-based scores.

Step 1: Upload Raw Ticket

Upload a raw ticket JSON file from your support system. The file should contain the ticket payload with interactions.

Step 2: Extract SLA Data

Step 3: Markdown Conversion

Step 4: Select Evaluation Rubric

Create New Rubric

Choose a rubric to evaluate the ticket against. You can use the built-in rubrics, upload a custom one, or create your own in the Rubric Editor.

Step 5: Run LLM Evaluation

Your API key is stored locally and never sent to our servers.

Format selection disabled when using custom prompt.

Judge Prompt

The prompt includes the rubric and ticket data. Customize instructions, output format, or evaluation criteria.

Evaluation Summary

Ticket:
Rubric:
Dimensions:
Prompt:
Format: