Agent Skills: Azure AI Text Analytics SDK for Python

Azure AI Text Analytics SDK for Python

Client library for Azure AI Language service NLP capabilities including sentiment, entities, key phrases, and more.

Installation

pip install azure-ai-textanalytics

Environment Variables

AZURE_LANGUAGE_ENDPOINT=https://<resource>.cognitiveservices.azure.com  # Required for all auth methods
AZURE_TOKEN_CREDENTIALS=prod # Required only if DefaultAzureCredential is used in production
AZURE_LANGUAGE_KEY=<your-api-key>  # Only required for the legacy API-key auth path below

Authentication & Lifecycle

🔑 Two rules apply to every code sample below:

1. Prefer DefaultAzureCredential. It works locally (Azure CLI / VS Code / Developer CLI) and in Azure (managed identity, workload identity) with no code change. Avoid connection strings, account/API keys — they bypass Entra audit and rotation.
   • Local dev: DefaultAzureCredential works as-is.
   • Production: set AZURE_TOKEN_CREDENTIALS=prod (or AZURE_TOKEN_CREDENTIALS=<specific_credential>) to constrain the credential chain to production-safe credentials.
2. Wrap every client in a context manager so HTTP transports, sockets, and token caches are released deterministically:
   • Sync: with <Client>(...) as client:
   • Async: async with <Client>(...) as client: and async with DefaultAzureCredential() as credential: (from azure.identity.aio)

Snippets may abbreviate this setup, but production code should always follow both rules.

import os
from azure.identity import DefaultAzureCredential, ManagedIdentityCredential
from azure.ai.textanalytics import TextAnalyticsClient

# Local dev: DefaultAzureCredential. Production: set AZURE_TOKEN_CREDENTIALS=prod or AZURE_TOKEN_CREDENTIALS=<specific_credential>
credential = DefaultAzureCredential(require_envvar=True)
# Or use a specific credential directly in production:
# See https://learn.microsoft.com/python/api/overview/azure/identity-readme?view=azure-python#credential-classes
# credential = ManagedIdentityCredential()

with TextAnalyticsClient(
    endpoint=os.environ["AZURE_LANGUAGE_ENDPOINT"],
    credential=credential,
) as client:
    languages = client.detect_language(["Hello, world!"])

Legacy: API Key (existing keyed deployments)

New code should use DefaultAzureCredential above. Use AzureKeyCredential only if you have an existing keyed deployment that hasn't been migrated to Entra ID yet — for example, regulated environments still completing their Entra rollout.

import os
from azure.core.credentials import AzureKeyCredential
from azure.ai.textanalytics import TextAnalyticsClient

with TextAnalyticsClient(
    endpoint=os.environ["AZURE_LANGUAGE_ENDPOINT"],
    credential=AzureKeyCredential(os.environ["AZURE_LANGUAGE_KEY"]),
) as client:
    languages = client.detect_language(["Hello, world!"])

Sentiment Analysis

documents = [
    "I had a wonderful trip to Seattle last week!",
    "The food was terrible and the service was slow."
]

result = client.analyze_sentiment(documents, show_opinion_mining=True)

for doc in result:
    if not doc.is_error:
        print(f"Sentiment: {doc.sentiment}")
        print(f"Scores: pos={doc.confidence_scores.positive:.2f}, "
              f"neg={doc.confidence_scores.negative:.2f}, "
              f"neu={doc.confidence_scores.neutral:.2f}")
        
        # Opinion mining (aspect-based sentiment)
        for sentence in doc.sentences:
            for opinion in sentence.mined_opinions:
                target = opinion.target
                print(f"  Target: '{target.text}' - {target.sentiment}")
                for assessment in opinion.assessments:
                    print(f"    Assessment: '{assessment.text}' - {assessment.sentiment}")

Entity Recognition

documents = ["Microsoft was founded by Bill Gates and Paul Allen in Albuquerque."]

result = client.recognize_entities(documents)

for doc in result:
    if not doc.is_error:
        for entity in doc.entities:
            print(f"Entity: {entity.text}")
            print(f"  Category: {entity.category}")
            print(f"  Subcategory: {entity.subcategory}")
            print(f"  Confidence: {entity.confidence_score:.2f}")

PII Detection

documents = ["My SSN is 123-45-6789 and my email is john@example.com"]

result = client.recognize_pii_entities(documents)

for doc in result:
    if not doc.is_error:
        print(f"Redacted: {doc.redacted_text}")
        for entity in doc.entities:
            print(f"PII: {entity.text} ({entity.category})")

Key Phrase Extraction

documents = ["Azure AI provides powerful machine learning capabilities for developers."]

result = client.extract_key_phrases(documents)

for doc in result:
    if not doc.is_error:
        print(f"Key phrases: {doc.key_phrases}")

Language Detection

documents = ["Ce document est en francais.", "This is written in English."]

result = client.detect_language(documents)

for doc in result:
    if not doc.is_error:
        print(f"Language: {doc.primary_language.name} ({doc.primary_language.iso6391_name})")
        print(f"Confidence: {doc.primary_language.confidence_score:.2f}")

Healthcare Text Analytics

documents = ["Patient has diabetes and was prescribed metformin 500mg twice daily."]

poller = client.begin_analyze_healthcare_entities(documents)
result = poller.result()

for doc in result:
    if not doc.is_error:
        for entity in doc.entities:
            print(f"Entity: {entity.text}")
            print(f"  Category: {entity.category}")
            print(f"  Normalized: {entity.normalized_text}")
            
            # Entity links (UMLS, etc.)
            for link in entity.data_sources:
                print(f"  Link: {link.name} - {link.entity_id}")

Multiple Analysis (Batch)

from azure.ai.textanalytics import (
    RecognizeEntitiesAction,
    ExtractKeyPhrasesAction,
    AnalyzeSentimentAction
)

documents = ["Microsoft announced new Azure AI features at Build conference."]

poller = client.begin_analyze_actions(
    documents,
    actions=[
        RecognizeEntitiesAction(),
        ExtractKeyPhrasesAction(),
        AnalyzeSentimentAction()
    ]
)

results = poller.result()
for doc_results in results:
    for result in doc_results:
        if result.kind == "EntityRecognition":
            print(f"Entities: {[e.text for e in result.entities]}")
        elif result.kind == "KeyPhraseExtraction":
            print(f"Key phrases: {result.key_phrases}")
        elif result.kind == "SentimentAnalysis":
            print(f"Sentiment: {result.sentiment}")

Async Client

from azure.ai.textanalytics.aio import TextAnalyticsClient
from azure.identity.aio import DefaultAzureCredential

async def analyze():
    async with DefaultAzureCredential() as credential:
        async with TextAnalyticsClient(
            endpoint=endpoint,
            credential=credential
        ) as client:
            result = await client.analyze_sentiment(documents)
            # Process results...

Client Types

| Client | Purpose | |--------|---------| | TextAnalyticsClient | All text analytics operations | | TextAnalyticsClient (aio) | Async version |

Available Operations

| Method | Description | |--------|-------------| | analyze_sentiment | Sentiment analysis with opinion mining | | recognize_entities | Named entity recognition | | recognize_pii_entities | PII detection and redaction | | recognize_linked_entities | Entity linking to Wikipedia | | extract_key_phrases | Key phrase extraction | | detect_language | Language detection | | begin_analyze_healthcare_entities | Healthcare NLP (long-running) | | begin_analyze_actions | Multiple analyses in batch |

Best Practices

1. Pick sync OR async and stay consistent. Do not mix azure.ai.textanalytics sync clients with azure.ai.textanalytics.aio async clients in the same call path. Choose one mode per module.
2. Always use context managers for clients and async credentials. Wrap every client in with TextAnalyticsClient(...) as client: (sync) or async with TextAnalyticsClient(...) as client: (async). For async DefaultAzureCredential from azure.identity.aio, also use async with credential: so tokens and transports are cleaned up.
3. Use batch operations for multiple documents (up to 10 per request)
4. Enable opinion mining for detailed aspect-based sentiment
5. Use async client for high-throughput scenarios
6. Handle document errors — results list may contain errors for some docs
7. Specify language when known to improve accuracy