Public trial evidence

50 rows, reproducible quality evidence.

This report documents a deterministic launch trial for WorxAI's synthetic human-trait generator. It is evidence about the generator's controls, not a claim that synthetic records are authentic human experiences or historical market data.

Rows requested

50

Rows accepted

50

Acceptance rate

100%

Quality score

0.960

What was tested

  • Deterministic seed: 42
  • Scenario: repairing a misunderstanding in a time-sensitive collaboration
  • Outputs: structured traces and multi-turn conversations
  • Controls: empathy, uncertainty, perspective-taking, social context, regulation, and emotional awareness
  • Rejected rows: 0; rejection reasons are retained in the API response

Quality gates

schema-valid
emotion-appraisal-consistent
causal-summary-present
diversity-checked
provenance-labeled

A row passes when required labels exist, numeric trait scores remain within bounds, and the appraisal, intent, and reasoning summary form a coherent labeled trace.

Representative sample

Synthetic output only. Use the API for the complete reproducible response.

Open JSON trial
{
  "trace": {
    "id": "human-trace-42-1",
    "synthetic": true,
    "scenario": "repairing a misunderstanding in a time-sensitive collaboration",
    "emotion": "calm",
    "emotionIntensity": 0.653,
    "appraisal": "uncertainty about outcome",
    "intent": "set a respectful boundary",
    "empathy": 0.83,
    "uncertainty": 0.742,
    "perspectiveTaking": 0.88,
    "emotionalRegulation": 0.82,
    "socialContext": "considers relationships, power, history, and likely impact",
    "reasoningSummary": "Notices uncertainty about outcome, considers another person’s likely perspective, and chooses to set a respectful boundary without overstating certainty.",
    "quality": {
      "score": 0.94,
      "method": "schema-completeness, range, causal-coherence, and diversity checks"
    }
  },
  "conversation": {
    "id": "human-conversation-42-1",
    "synthetic": true,
    "scenario": "repairing a misunderstanding in a time-sensitive collaboration",
    "turns": [
      {
        "speaker": "person_a",
        "emotion": "calm",
        "text": "I’m feeling calm because I’m not sure what will happen next."
      },
      {
        "speaker": "person_b",
        "emotion": "attentive",
        "text": "That makes sense. I hear the uncertainty, and I want to understand what would feel fair and useful."
      },
      {
        "speaker": "person_a",
        "emotion": "cautiously hopeful",
        "text": "I’d like to set a respectful boundary, while keeping room to revise if we learn more."
      }
    ],
    "labels": {
      "empathy": 0.9,
      "uncertainty": 0.78,
      "perspectiveTaking": 0.88,
      "repairOrientation": 0.82
    },
    "quality": {
      "score": 0.94,
      "method": "dialogue-coherence, empathy, uncertainty, and safety checks"
    }
  }
}

Buy datasets

Pricing, licensing, and buyer FAQ.

OpenAPI

Machine-readable endpoints and payment details.

Manifest

Live catalog and x402 resource metadata.