Getting Started¶

Install¶

Core only:
```
pip install aiobs
```
With OpenAI support:
```
pip install aiobs[openai]
```
With Gemini support:
```
pip install aiobs[gemini]
```

With all providers and examples:

pip install aiobs[openai,gemini,example]

Get Your API Key¶

An API key is required to use aiobs. Get your free API key from:

👉 https://neuralis-in.github.io/shepherd/api-keys

Once you have your API key, set it as an environment variable:

export AIOBS_API_KEY=aiobs_sk_your_key_here

Or add it to your .env file:

AIOBS_API_KEY=aiobs_sk_your_key_here

You can also pass the API key directly when starting a session:

observer.observe(api_key="aiobs_sk_your_key_here")

You can add labels for filtering in enterprise dashboards:

observer.observe(
    api_key="aiobs_sk_your_key_here",
    labels={
        "environment": "production",
        "team": "ml-platform",
    }
)

Configure Provider Credentials¶

Create a .env at your project root with provider credentials:

OpenAI:

OPENAI_API_KEY=your_key_here
# Optional
OPENAI_MODEL=gpt-4o-mini

Google Gemini:

GOOGLE_API_KEY=your_key_here
# Or use service account credentials
GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account.json

Hello, Observability¶

Minimal example with OpenAI:

from aiobs import observer
from openai import OpenAI

observer.observe()

client = OpenAI()
response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Hello!"}]
)

observer.end()
observer.flush()  # writes llm_observability.json

Minimal example with Gemini:

from aiobs import observer
from google import genai

observer.observe()

client = genai.Client()
response = client.models.generate_content(
    model="gemini-2.0-flash",
    contents="Hello!"
)

observer.end()
observer.flush()  # writes llm_observability.json