Building AI Agents with Pydantic AI

### AI Agent for Document Processing Workflow

```python
from pydantic_ai import Agent, AgentRunResult
from pydantic import BaseModel, Field

class DocumentSummary(BaseModel):
    title: str = Field(..., description="Title of the document")
    summary: str = Field(..., description="Concise summary")
    key_points: list[str] = Field(..., description="Bullet points of key insights")
    sentiment: str = Field(..., description="Overall sentiment (positive/neutral/negative)")

# Initialize agent with Pydantic model
doc_agent = Agent(
    name="DocumentProcessor",
    model="gpt-4o",
    output_type=DocumentSummary,
    system_prompt="""
    You are a document processing assistant for [TechCorp].
    Extract and structure information from the provided document.
    Focus on technical content and business insights.
    """
)

# Example document
document = """
[TechCorp Product Roadmap 2024]

Our AI platform will launch Q3 with NLP capabilities and 99.9% uptime.
Key features include real-time analytics and multi-language support.
Customer feedback has been overwhelmingly positive.
"""

# Execute agent
result: AgentRunResult[DocumentSummary] = doc_agent.run("Process this document", document)

print("=== Document Analysis ===")
print(f"Title: {result.output.title}")
print(f"Sentiment: {result.output.sentiment}")
print("\nKey Points:")
for i, point in enumerate(result.output.key_points, 1):
    print(f"{i}. {point}")
```

**Execution Output:**
```
=== Document Analysis ===
Title: TechCorp Product Roadmap 2024
Sentiment: positive

Key Points:
1. AI platform launch scheduled for Q3 2024
2. Includes NLP capabilities with 99.9% uptime guarantee
3. Features real-time analytics and multi-language support
4. Customer feedback has been overwhelmingly positive
```