Add AzureOpenAI as a model provider to DraftValidation

[howardbaik]

Summary

• pointblank/draft.py: new azure-openai branch in __post_init__ after the bedrock block, mirroring the _utils_ai.py pattern but using self.api_key and the draft-specific system prompt. Docstring updated in three places: class summary, model parameter description, and the "Constructing the model Argument" block (now lists five providers and explains the :deployment_id convention plus required env vars).
• pointblank/draft.py:254: fixed an unrelated latent Windows bug where api-docs.txt was being read with the platform default encoding (cp1252 on Windows), which crashed on non-latin1 bytes. Now uses encoding="utf-8". This only surfaced because the new tests reach further into __post_init__ than the old ones.
• user_guide/02-advanced-validation/04-draft-validation.qmd: added Azure OpenAI to the supported-providers list, to the .env sample, and to the model-specification examples.
• tests/test_draft.py: two new tests covering the missing-endpoint and missing-api-version ValueError paths, following the existing monkeypatch/importorskip style from test__utils_ai.py.

Example Usage

Reproducible Test Code

import pointblank as pb
import os
from dotenv import load_dotenv

load_dotenv(override=True)

api_key = os.getenv("AZURE_OPENAI_API_KEY")

data = pb.load_dataset(dataset="global_sales", tbl_type="polars")

# Generate a validation plan
pb.DraftValidation(
    data=data,
    model="azure-openai:gpt-5.4",
    api_key=api_key
)

Expected Output

import pointblank as pb

# Define schema based on column names and dtypes
schema = pb.Schema(
    columns=[
        ("product_id", "String"),
        ("product_category", "String"),
        ("customer_id", "String"),
        ("customer_segment", "String"),
        ("region", "String"),
        ("country", "String"),
        ("city", "String"),
        ("timestamp", "Datetime(time_unit='us', time_zone=None)"),
        ("quarter", "String"),
        ("month", "Int64"),
        ("year", "Int64"),
        ("price", "Float64"),
        ("quantity", "Int64"),
        ("status", "String"),
        ("email", "String"),
        ("revenue", "Float64"),
        ("tax", "Float64"),
        ("total", "Float64"),
        ("payment_method", "String"),
        ("sales_channel", "String"),
    ]
)

# The validation plan
validation = (
    pb.Validate(
        data=your_data,  # Replace your_data with the actual data variable
        label="Draft Validation",
        thresholds=pb.Thresholds(warning=0.10, error=0.25, critical=0.35)
    )
    .col_schema_match(schema=schema)
    .col_count_match(count=20)
    .row_count_match(count=50000)
    .rows_distinct()
    .col_vals_not_null(
        columns=[
            "product_category",
            "customer_segment",
            "region",
            "country",
            "month",
            "year",
            "price",
            "quantity",
            "status",
            "email",
            "revenue",
            "tax",
            "total",
            "payment_method",
            "sales_channel",
        ]
    )
    .col_vals_between(
        columns="month",
        left=1,
        right=12,
        na_pass=True
    )
    .col_vals_between(
        columns="year",
        left=2021,
        right=2023,
        na_pass=True
    )
    .col_vals_gt(
        columns="price",
        value=0,
        na_pass=False
    )
    .col_vals_gt(
        columns="quantity",
        value=0,
        na_pass=False
    )
    .col_vals_in_set(
        columns="status",
        set=[
            "pending",
            "processing",
            "shipped",
            "delivered",
            "returned",
            "cancelled",
        ]
    )
    .col_vals_within_spec(
        columns="email",
        spec="email"
    )
    .col_vals_gt(
        columns="revenue",
        value=0,
        na_pass=False
    )
    .col_vals_ge(
        columns="tax",
        value=0,
        na_pass=False
    )
    .col_vals_ge(
        columns="total",
        value=0,
        na_pass=False
    )
    .col_vals_in_set(
        columns="quarter",
        set=[
            "2021-Q1",
            "2021-Q2",
            "2021-Q3",
            "2021-Q4",
            "2022-Q1",
            "2022-Q2",
            "2022-Q3",
            "2022-Q4",
            "2023-Q1",
            "2023-Q2",
            "2023-Q3",
            "2023-Q4",
        ]
    )
    .col_vals_in_set(
        columns="region",
        set=[
            "North America",
            "Europe",
            "Asia Pacific",
        ]
    )
    .col_vals_in_set(
        columns="payment_method",
        set=[
            "Credit Card",
            "PayPal",
            "Bank Transfer",
            "Apple Pay",
            "Google Pay",
        ]
    )
    .col_vals_in_set(
        columns="sales_channel",
        set=[
            "Online",
            "Retail",
            "Distributor",
            "Partner",
            "Phone",
        ]
    )
    .interrogate()
)

validation

Note

Make sure OPENAI_API_VERSION is set to 2025-03-01-preview or later — the Azure OpenAI Responses API is only available on these versions.

Related GitHub Issues and PRs

Fixes: #386

Checklist

• [ x] I understand and agree to the Code of Conduct.
• [ x] I have followed the Style Guide for Python Code as best as possible for the submitted code.
• [ x] I have added pytest unit tests for any new functionality.

[rich-iannone]

LGTM, thanks!