Installation and setup

Standard setup (recommended)

Use uv for project management and install the SDK from the private Backbase Artifactory.

1. Configure `pyproject.toml`

Add the Backbase Artifactory repository to your pyproject.toml. This tells uv where to find the bb-ai-sdk package.

pyproject.toml

[[tool.uv.index]]
name = "backbase"
url = "https://repo.backbase.com/artifactory/api/pypi/pypi-local/simple"

2. Set credentials

Authenticate with Artifactory using environment variables. Add these to your .env file (and ensure .env is in .gitignore):

.env

# Artifactory credentials
export UV_INDEX_BACKBASE_USERNAME=your_username
export UV_INDEX_BACKBASE_PASSWORD=your_api_key_or_password

The environment variable format is UV_INDEX_<NAME>_USERNAME, where <NAME> matches the name you defined in pyproject.toml (uppercased). You must export them in .env file

3. Install the SDK

Load your environment variables and add the package:

# Load env vars
source .env

# Install SDK (current platform release: 0.1.9)
uv add "bb-ai-sdk==0.1.9" --index backbase

Observability extras (recommended for agents)

Install the instrumentation bundles you need. These pull OpenInference + OTel helpers; they do not install Agno/LangChain themselves.

# FastAPI HTTP service + Agno tracing (typical agent API)
uv add "bb-ai-sdk[instrument-fastapi,instrument-agno]==0.1.9" --index backbase

# LangChain or LangGraph tracing
uv add "bb-ai-sdk[instrument-langchain]==0.1.9" --index backbase
# or
uv add "bb-ai-sdk[instrument-langgraph]==0.1.9" --index backbase

`framework=` in code	Extra
`"agno"`	`[instrument-agno]` (legacy alias: `[agno]`)
`"langchain"`	`[instrument-langchain]`
`"langgraph"`	`[instrument-langgraph]`
FastAPI route spans	`[instrument-fastapi]`

Guardrails: uv add "bb-ai-sdk[guardrails]==0.1.9" --index backbase

Verification

Verify the installation was successful (expect 0.1.9):

uv run python -c "import bb_ai_sdk; print(bb_ai_sdk.__version__)"

Overview

Architecture and technology

Agent Development Lifecycle

Get started

Starter kits

BB AI SDK

MCP support

CI/CD workflows

Standard setup (recommended)

1. Configure `pyproject.toml`

2. Set credentials

3. Install the SDK

Observability extras (recommended for agents)

Verification

​Standard setup (recommended)

​1. Configure pyproject.toml

​2. Set credentials

​3. Install the SDK

​Observability extras (recommended for agents)

​Verification

Standard setup (recommended)

1. Configure `pyproject.toml`

2. Set credentials

3. Install the SDK

Observability extras (recommended for agents)

Verification