An API that judges customer support tickets — returning choices, scores or probabilities so your thresholds decide.
judged.systems homepage
1 / 2Overview
What is judged.systems?
A judgment engine for customer support. Bring your own helpdesk.
judged.systems is a judgment engine for customer support: support systems call it when a ticket needs a judgment, while your helpdesk stays the system of record. You send one ticket; a pack defines the questions (queue, urgency, refund request, policy risk); what comes back is a choice with probabilities, a rubric score, or the probability a statement is true.
Your own thresholds turn that evidence into accept or review — the model never makes the final call, and failed calls stay in review. Tickets are redacted first (PII removed before storage), and evaluations are stored as they happened: a label on a miss does not rewrite the evaluation.
Four integration paths: an app playground, REST POST /api/judge, MCP, and inbound webhooks. Free to use per the Product Hunt launch; no pricing page on the site yet. Documentation at judged.systems/docs.
Platforms and languages
judged.systems Availability
Platforms
Languages
Capabilities
judged.systems Key Features
Pack-based questions
Define packs asking queue, urgency, refund request, and policy risk questions per ticket.
Typed evidence
Returns a choice with probabilities, a rubric score, or a statement probability; the model never makes the final call.
Threshold-driven decisions
User thresholds map evidence to accept/review/reject; failed calls stay in review.
Four integration paths
App playground, REST POST /api/judge, MCP, and inbound webhooks.
Redaction plus immutable audit
PII redacted before storage; evaluations stored as they happened; labels on misses do not rewrite evaluations.
Best for
Who uses judged.systems?
Customer support teams and helpdesk operators
Get structured judgments on support tickets to route, prioritize and review faster
Plans and access
judged.systems Pricing
Free
Free; no pricing page on site
Common questions
judged.systems FAQs
How does a ticket get judged?
Send one ticket via REST, webhook, or MCP. The engine asks the pack's questions and returns typed evidence.
What comes back from a judgment?
A choice with probabilities, a rubric score, or the probability a statement is true. Your thresholds turn it into accept or review.
Does the model make the final call?
No. The model returns evidence only; your thresholds decide, and a failed call stays in review.
Is my ticket data private?
Tickets are redacted first; evaluations are stored as they happened, and labeling a miss leaves them untouched.
Which integrations exist?
A REST API, MCP, inbound webhooks, and an app playground.
Reviews
See what the community thinks and share your experience.
—
Based on 0 ratings
Rating distribution
Community reviews
No reviews
No written reviews yet.
