The Oracle AI Optimizer and Toolkit (the AI Optimizer) provides a streamlined environment where developers and data scientists can explore the potential of Generative Artificial Intelligence (GenAI).

By integrating Oracle AI Database Vector Search and SQLcl MCP with familiar Open Source Software (OSS), the AI Optimizer enables users to enhance existing Language Models through Retrieval Augmented Generation (RAG) and Natural Language to SQL (NL2SQL).

This method significantly improves the accuracy of AI models, helping to avoid common issues such as knowledge cut-off and hallucinations.

Familiar OSS components^*

• FastAPI (REST API)
• FastMCP (tools)
• LangGraph + AgentSpec (orchestration)
• Docling (parsing)
• OpenTelemetry (observability)
• LiteLLM (models)
• Giskard (evaluation)

* Project names and logos are shown for identification purposes only. Unless expressly stated, their use does not imply affiliation with, sponsorship by, or endorsement by Oracle or the respective project owners.

Features

The AI Optimizer brings together the capabilities you need to build, test, and refine GenAI solutions on the Oracle AI Database:

• Configuring Embedding and Language Models
• Experimenting with Language Model Parameters
• Splitting and Embedding Documentation
• Modifying System Prompts (Prompt Engineering)
• Enforcing Deep Data Security
• Testbed for auto-generated or existing Q&A datasets

Getting Started

The AI Optimizer is available to install in your own environment, which may be a developer’s desktop, on-premises data center environment, or a cloud provider. It can be run either on bare-metal, within a container, or in a Kubernetes Cluster.

Prerequisites

• Python 3.11 (for running Bare-Metal)
• Container Runtime e.g. docker/podman (for running in a Container)
• Access to a Chat Model, and an Embedding Model for RAG:
  • API Keys for Third-Party Models
  • On-Premises Models*
• Oracle AI Database incl. Oracle AI Database Free

_{*Oracle recommends running On-Premises Models on hardware with GPUs. For more information, please review the AI Optimizer documentation.}

Available deployment methods:

• Bare-Metal Installation
• Container Installation
• Oracle Cloud Infrastructure

Bare-Metal Installation

To run the application on bare-metal, download the latest release:

• TAR.GZ archive
• ZIP archive

1. Uncompress the release in a new directory. For example:

   mkdir ai-optimizer
   tar zxf ai-optimizer-src.tar.gz -C ai-optimizer
   
   cd ai-optimizer

2. Create and activate a Python Virtual Environment:

   cd ai-optimizer
   python3.11 -m venv .venv
   source .venv/bin/activate
   pip3.11 install --upgrade pip wheel uv

3. Install the Python modules:

   uv pip install -e ".[all]"

4. (Optional) Create an environment file to pre-configure the application:

   cp src/.env.example src/.env.dev

   Edit src/.env.dev as needed. See Environment Variables for details.

5. Start the application:

   python src/entrypoint.py client

6. Navigate to http://localhost:8501.

7. Configure the AI Optimizer.

Container Installation

To run the application in a container, download the latest release:

• TAR.GZ archive
• ZIP archive

1. Uncompress the release in a new directory. For example:

   mkdir ai-optimizer
   tar zxf ai-optimizer-src.tar.gz -C ai-optimizer
   
   cd ai-optimizer

2. Build the ai-optimizer-aio image.

   Note: MacOS Silicon users may need to specify --arch amd64

   podman build -f src/Dockerfile -t ai-optimizer-aio .

3. Start the Container:

   podman run -p 8501:8501 -it --rm ai-optimizer-aio

4. Navigate to http://localhost:8501.

5. Configure the AI Optimizer.

Oracle Cloud Infrastructure

The AI Optimizer can easily be deployed using Infrastructure as Code (IaC) into Oracle Cloud Infrastructure (OCI).

OCI deployment options include:

• An Always-Free Installation
• Simple Virtual Machine Installation
• Advanced Oracle Kubernetes Engine

To get started, review the IaC documentation.