Coming Soon

Developer Documentation

Integrate Mizaan's scoring engine into your workflow with a simple REST API. Provider-agnostic, feedback-aware, and built for scale.

Quick Start

Python
from mizaan import ScoringEngine
from mizaan.providers import OllamaProvider

# Initialize with local LLM — zero cloud dependency
engine = ScoringEngine(
    provider=OllamaProvider(model="mistral"),
    rubric="sales-performance-v2"
)

# Score a work item
result = engine.score(
    item="90-min discovery call with Fortune 500 CFO",
    context={"type": "sales_call", "duration": 90}
)

print(result.score)       # 8.5
print(result.reasoning)   # "Extended engagement with..."
print(result.confidence)  # 0.87
print(result.dimensions)  # {"impact": 9, "effort": 8, ...}

API Reference

POST
/api/v1/score

Score a work item with feedback-aware context

Request Body

{
  "work_item_id": "wi_001",
  "work_item_type": "sales_call",
  "description": "90-minute discovery call with Fortune 500 CFO...",
  "context": { "duration": 90, "outcome": "follow_up_scheduled" }
}

Response

{
  "record_id": "fb_001",
  "score": 8.5,
  "reasoning": "Extended engagement with C-suite decision-maker...",
  "confidence": 0.87,
  "dimensions": { "impact": 9, "effort": 8, "quality": 9 },
  "auditor_decision": "APPROVED",
  "similar_cases_used": 3
}
POST
/api/v1/feedback

Submit human feedback on a scoring result

Request Body

{
  "record_id": "fb_001",
  "decision": "OVERRIDDEN",
  "adjusted_score": 9.0,
  "reasoning": "Fortune 500 CFO calls are extremely valuable..."
}

Response

{
  "status": "saved",
  "record_id": "fb_001"
}
GET
/api/v1/patterns

Get detected override patterns for rubric improvement

Response

{
  "patterns": [
    {
      "pattern": "executive_meetings_underscored",
      "description": "AI consistently scores C-suite meetings 2+ points below human expectation",
      "occurrences": 12,
      "suggested_adjustment": "Increase weight for seniority_of_contact dimension"
    }
  ]
}

Built With

Python
FastAPI
LangChain
PostgreSQL
pgvector
Redis
Docker
Ollama