دليل هندسي: Freelance AI Programmer: 100% Deterministic LLM Structured Outputs with Pydantic. ألديان فازريهادي (Aldian Fazrihady) مبرمج مستقل أول ومهندس ذكاء اصطناعي (Python, Django, FastAPI, React, Next.js, TypeScript, Golang).

Stop broken JSON parsing in production AI apps by hiring freelance AI programmer Aldian Fazrihady to architect constrained decoding and Pydantic validation pipelines.

from typing import Literal, List
from pydantic import BaseModel, Field, field_validator

class LineItemExtraction(BaseModel):
    sku: str = Field(min_length=2, max_length=64)
    quantity: int = Field(gt=0)
    unit_price_cents: int = Field(ge=0)

class InvoiceExtractionResult(BaseModel):
    vendor_name: str
    currency: Literal["USD", "EUR", "GBP", "SGD", "AUD", "IDR"]
    line_items: List[LineItemExtraction]
    total_cents: int = Field(ge=0)

    @field_validator("total_cents")
    @classmethod
    def verify_math_consistency(cls, v: int, info) -> int:
        items = info.data.get("line_items", [])
        expected = sum(i.quantity * i.unit_price_cents for i in items)
        if items and v != expected:
            raise ValueError(f"Line items sum ({expected}) != total_cents ({v})")
        return v

توظيف عبر Upwork | دليل 455+ مقال هندسي