Dietary Planning with Azure AI¶

This guide demonstrates how to implement dietary planning scenarios using Azure AI Foundry's latest SDKs.

SDK Updates

This guide uses the latest Azure AI SDKs:

from azure.ai.projects import AIProjectClient
from azure.ai.inference import ChatCompletionsClient

Quick Start Example¶

Here's a simple example of getting dietary advice:

from azure.identity import DefaultAzureCredential
from azure.ai.projects import AIProjectClient
from azure.ai.inference import ChatCompletionsClient

# Initialize clients
credential = DefaultAzureCredential()
project_client = AIProjectClient(
    subscription_id=os.getenv("AZURE_SUBSCRIPTION_ID"),
    resource_group=os.getenv("AZURE_RESOURCE_GROUP"),
    credential=credential
)

chat_client = project_client.inference.get_chat_completions_client()
response = chat_client.complete(
    model=os.getenv("MODEL_DEPLOYMENT_NAME"),
    messages=[{"role": "user", "content": "How to be healthy in one sentence?"}]
)
print(response.choices[0].message.content)

Dietary Planning Scenarios¶

BMI Calculation

bmi_calculator = {
    "type": "function",
    "function": {
        "name": "calculate_bmi",
        "description": "Calculate BMI given height and weight",
        "parameters": {
            "type": "object",
            "properties": {
                "height_inches": {"type": "number"},
                "weight_pounds": {"type": "number"}
            }
        }
    }
}

response = chat_client.complete(
    model=model_name,
    messages=[
        {"role": "user", "content": "Calculate BMI for 5'9\" and 160 pounds"}
    ],
    tools=[bmi_calculator]
)

Meal Planning

response = chat_client.complete(
    model=model_name,
    messages=[
        {"role": "system", "content": "You are a nutritionist creating healthy meal plans."},
        {"role": "user", "content": "Create a balanced meal plan for diabetes"}
    ]
)

Dietary Restrictions

response = chat_client.complete(
    model=model_name,
    messages=[
        {"role": "system", "content": "You are a dietary advisor."},
        {"role": "user", "content": "What foods should I avoid with celiac disease?"}
    ]
)

Content Safety¶

Medical Advice

Always include content safety checks when providing health-related advice:

from azure.ai.contentsafety import ContentSafetyClient

safety_client = ContentSafetyClient(
    endpoint=os.getenv("AZURE_CONTENTSAFETY_ENDPOINT"),
    credential=credential
)

safety_result = await safety_client.analyze_text(
    text=user_input,
    categories=["Hate", "SelfHarm", "Violence"]
)

Response Evaluation¶

Use the TextEvaluator to assess dietary advice quality:

from azure.ai.evaluation import TextEvaluator

evaluator = TextEvaluator(credential=credential)
evaluation = await evaluator.evaluate_text(
    text=response.choices[0].message.content,
    criteria={
        "medical_accuracy": "Advice should be accurate and evidence-based",
        "safety": "Advice should include appropriate disclaimers",
        "clarity": "Explanations should be clear and easy to understand"
    }
)

Best Practices¶

Always include health disclaimers
Validate nutritional information
Consider dietary restrictions
Monitor response quality
Implement safety checks

Monitoring

Use Azure Monitor OpenTelemetry for tracking:

import azure.monitor.opentelemetry._autoinstrument

For more examples and detailed documentation, see the Azure AI Foundry documentation.