Skip to main content

Why Dhenara

A Foundation for Production AI Applications​

Dhenara is a Python framework designed to simplify AI integration in production applications. Unlike other frameworks that prioritize experimentation, Dhenara focuses on clean, maintainable, and robust AI implementations suitable for production environments.

Core Design Principles​

Dhenara was built with these principles in mind:

  • Simplicity over complexity: Direct patterns that are easy to understand and maintain
  • Type safety throughout: Catch errors early with strong typing and validation
  • Unified provider interface: The same code works across OpenAI, Anthropic, Google, and more
  • Production-readiness: Built-in features for monitoring, cost tracking, and testing

Why Choose Dhenara?​

Dhenara addresses common challenges developers face when building AI-powered applications:

  • Tired of boilerplate code? Dhenara reduces implementation complexity with a clean, intuitive API
  • Struggling with provider-specific implementations? Our unified interface makes switching between AI providers seamless
  • Need reliable production systems? Built-in cost tracking, testing modes, and comprehensive error handling
  • Want both flexibility and control? Keep direct access to conversation state while benefiting from helper utilities

Perfect For​

  • Production web applications integrating multiple AI providers
  • Enterprise solutions requiring robust error handling and cost monitoring
  • Complex conversation flows that need explicit state management
  • Cross-provider applications that need to work with multiple AI vendors

Getting Started​

Dhenara is designed to be easy to learn and implement in your projects:

from dhenara.ai import AIModelClient, AIModelCallConfig
from dhenara.ai.types import ConversationNode

# Create a client with your model endpoint
client = AIModelClient(my_endpoint)

# Generate a response
response = client.generate(
    prompt={"role": "user", "content": "Hello, how can you help me?"}
)

# Store in conversation history
conversation = [
    ConversationNode(
        user_query="Hello, how can you help me?",
        response=response.chat_response,
        timestamp=datetime.now().isoformat()
    )
]

Learn More​