Lucas AI Integration API

A central API for attaching AI chatbots to Lucas’s systems without re-implementing logic every time. This project centralizes AI chatbot infrastructure across multiple projects, allowing systems to connect and inherit a shared configuration.

Features

• Centralized API key management: Manage access to AI services in one place.
• Model selection and version control: Consistent AI behavior across all attached systems.
• Shared RAG (Retrieval-Augmented Generation) pipelines: Uses a shared knowledge store for contextual responses.
• Consistent request and response formats: Simplifies integration for various front-end and back-end systems.

Tech Stack

• Python (Flask): Backend API framework.
• Google Gemini API: Powering the AI responses and embeddings.
• SQLite: Stores knowledge chunks and their corresponding embeddings for RAG.
• Tailwind CSS: Used for the minimal landing page.

Getting Started

Prerequisites

• Python 3.13+
• A Google Gemini API Key

Installation

1. Clone the repository:

   git clone https://github.com/RLucasLR/ai-api.git
   cd ai-api

2. Create and activate a virtual environment:

   python -m venv .venv
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Set up environment variables in a .env file:

   GOOGLE_API_KEY=your_gemini_api_key
   PORTFOLIO_CHAT_API_KEY=your_custom_api_key_for_accessing_this_api

Database Initialization

The project uses a SQLite database (rag.db) to store embeddings. The table is automatically created when you run main.py. To populate it with data, you can use the embedding script.

Note: Before running the script, ensure you have the correct source file you want to chunk. You can change the target file by modifying the SOURCE_FILE constant at the top of scripts/embedding.py.

python scripts/embedding.py

Running the API

python main.py

The API will be available at http://127.0.0.1:5000.

API Documentation

POST /portfolio-chat

Interact with the AI assistant.

Request Body:

{
  "api_key": "your_portfolio_chat_api_key",
  "prompt": "Hello, who are you?",
  "previous_chats": [
    {"role": "user", "content": "Hi"},
    {"role": "model", "content": "Hello! I am Lucas's AI assistant."}
  ]
}

Response:

{
  "chat": "I am an AI assistant embedded on the personal portfolio website of Lucas Reddington..."
}

Project Structure

• main.py: Entry point of the Flask application and API routes.
• services/ai.py: Core AI logic, RAG implementation, and Gemini integration.
• scripts/embedding.py: Script for generating and storing embeddings in the database.
• templates/: HTML templates for the landing page and testing.
• rag.db: SQLite database for storing portfolio information and embeddings.