OpenAI Backend

The openai backend integrates with OpenAI’s GPT models and Azure OpenAI deployments.

Configuration

OpenAI API

Set the OPENAI_API_KEY environment variable:

export OPENAI_API_KEY="sk-..."  # pragma: allowlist secret

Or store it in ~/.config/kanoa/.env:

mkdir -p ~/.config/kanoa
echo "OPENAI_API_KEY=your-api-key-here" >> ~/.config/kanoa/.env

Initialize with:

from kanoa import AnalyticsInterpreter

interpreter = AnalyticsInterpreter(
    backend="openai",
    model="gpt-4o"  # or 'gpt-4-turbo', 'gpt-3.5-turbo'
)

Azure OpenAI

For Azure OpenAI deployments:

interpreter = AnalyticsInterpreter(
    backend="openai",
    api_base="https://your-resource.openai.azure.com/openai/deployments/your-deployment",
    api_key="your-azure-key",  # pragma: allowlist secret
    api_version="2024-02-01"
)

Or configure via environment variables:

# ~/.config/kanoa/.env
AZURE_OPENAI_API_KEY=your-azure-key
AZURE_OPENAI_ENDPOINT=https://your-resource.openai.azure.com
AZURE_OPENAI_DEPLOYMENT=your-deployment-name

Supported Models

Recommended Models

gpt-4o: Latest and most capable multimodal model
gpt-4-turbo: Fast, high-quality responses with vision support
gpt-4: Most capable for complex reasoning tasks
gpt-3.5-turbo: Cost-effective for simpler tasks

All models support vision capabilities (can process matplotlib figures).

Features

Vision Capabilities

All GPT-4 models support multimodal inputs. kanoa automatically converts matplotlib figures to images and sends them to the model.

Knowledge Base

The OpenAI backend supports both Text and PDF knowledge bases:

# Text knowledge base
interpreter = interpreter.with_kb(kb_path="data/docs")  # Auto-detects file types

# PDF knowledge base

Cost Tracking

kanoa automatically tracks token usage and costs for OpenAI models:

result = interpreter.interpret(fig=plt.gcf(), context="Analysis")
print(f"Cost: ${result.usage.total_cost:.4f}")
print(f"Input tokens: {result.usage.prompt_tokens}")
print(f"Output tokens: {result.usage.completion_tokens}")

Pricing is based on current OpenAI API rates and updated regularly.

Advanced Configuration

Custom Parameters

interpreter = AnalyticsInterpreter(
    backend="openai",
    model="gpt-4o",
    temperature=0.7,
    max_tokens=2048,
    top_p=0.9
)

Streaming Responses

# Streaming is not yet supported in the current version
# Coming in a future release