Select AI for Python

Select AI for Python enables you to ask questions of your database data using natural language (text-to-SQL), get generative AI responses using your trusted content (retrieval augmented generation), generate synthetic data using large language models, and other features – all from Python. With the general availability of Select AI Python, Python developers have access to the functionality of Select AI on Oracle Autonomous Database.

Select AI for Python enables you to leverage the broader Python ecosystem in combination with generative AI and database functionality - bridging the gap between the DBMS_CLOUD_AI PL/SQL package and Python's rich ecosystem. It provides intuitive objects and methods for AI model interaction.

Table of Contents

• Installation
• Documentation
• Getting Started
  • Async Example
• Command Line Interface
  • Chat
  • A2A Server
    • Cloud Run
• Samples
• Help
• Contributing
• Security
• License

Installation

Install the Python package:

python3 -m pip install select_ai

Install the optional command line interface:

python3 -m pip install 'select_ai[cli]'

The CLI extra includes A2A server support.

Documentation

See Select AI for Python documentation

Getting Started

import select_ai

user = "<your_select_ai_user>"
password = "<your_select_ai_password>"
dsn = "<your_select_ai_db_connect_string>"

select_ai.connect(user=user, password=password, dsn=dsn)
profile = select_ai.Profile(profile_name="oci_ai_profile")
# run_sql returns a pandas dataframe
df = profile.run_sql(prompt="How many promotions?")
print(df.columns)
print(df)

Async Example

import asyncio

import select_ai

user = "<your_select_ai_user>"
password = "<your_select_ai_password>"
dsn = "<your_select_ai_db_connect_string>"

# This example shows how to asynchronously run sql
async def main():
    await select_ai.async_connect(user=user, password=password, dsn=dsn)
    async_profile = await select_ai.AsyncProfile(
        profile_name="async_oci_ai_profile",
    )
    # run_sql returns a pandas df
    df = await async_profile.run_sql("How many promotions?")
    print(df)

asyncio.run(main())

Command Line Interface

The optional select-ai command provides interactive chat, SQL, profile management, and A2A server tools for Select AI:

Chat

select-ai chat --profile OCI_AI_PROFILE

A2A Server

Expose one Oracle Database AI agent team as an A2A JSON-RPC HTTP server:

select-ai a2a serve --team SALES_ANALYST --port 8000

The command obtains database connection settings from its options or the SELECT_AI_* environment variables. Its Agent Card is available at /.well-known/agent-card.json, and its JSON-RPC endpoint is /a2a/jsonrpc/. Set --public-url when the server is behind a proxy or load balancer so that clients receive its externally reachable URL.

For Autonomous Database mTLS, also set SELECT_AI_WALLET_LOCATION to the directory containing the unzipped wallet and set SELECT_AI_WALLET_PASSWORD. The CLI passes both values to the Select AI SDK as wallet_location and wallet_password.

The server accepts both A2A 1.x and the A2A v0.3 JSON-RPC streaming protocol for compatibility with Gemini Enterprise.

See the A2A user guide for dynamic sessions, the A2UI connection flow, persistent task state, task polling and cancellation, wallet configuration, and Google Cloud deployment modes.

Generate the A2A v0.3 Agent Card to paste into Gemini Enterprise after the service has a public URL:

select-ai a2a agent-card \
  --team ORACLE_AI_DATABASE_AGENT \
  --public-url https://YOUR-SERVICE.run.app

Cloud Run

Deploy the A2A server to Cloud Run using the instructions in gcloud/README.md.

Samples

For in-depth examples, see the /samples directory.

Help

Questions can be asked in GitHub Discussions.

Problem reports can be raised in GitHub Issues.

Contributing

This project welcomes contributions from the community. Before submitting a pull request, please review our contribution guide

Security

Please consult the security guide for our responsible security vulnerability disclosure process

License

Copyright (c) 2025, 2026 Oracle and/or its affiliates.

Released under the Universal Permissive License v1.0 as shown at https://oss.oracle.com/licenses/upl/.