Managing Multiple Chat Sessions with OpenAI in PHP

Welcome to the next step in your journey of creating a chatbot with OpenAI! In the previous lessons, you learned how to send messages to OpenAI's language model, explored model parameters, maintained conversation history, and personalized AI behavior with system prompts. Now, we will focus on managing multiple chat sessions. This is crucial for applications where you need to handle several conversations simultaneously, such as customer service chatbots. By the end of this lesson, you will be able to create and manage multiple chat sessions using OpenAI's API, setting the stage for more complex interactions.

This unit uses global arrays like $chatSessions to demonstrate how separate conversations can be tracked. That is fine for a teaching example, but production code should usually encapsulate session state in a class or persistent store and avoid depending on mutable global state.

For the same reason, uniqid() is acceptable here as a simple demo tool, but production systems should prefer stronger identifiers such as:

$chatId = bin2hex(random_bytes(16));

Also avoid logging full conversation histories in production unless you have a clear need and appropriate safeguards for sensitive user data.

In a chatbot application, each conversation should be treated as a separate session. To achieve this, we use unique identifiers for each chat session. This ensures that messages and responses are correctly associated with their respective sessions. In our code example, we use PHP's uniqid() function to generate a unique identifier for each chat session. When a new chat session is created, a unique chat_id is generated, and an empty conversation history is initialized.

require 'vendor/autoload.php';

// Fetch the api key
$apiKey = $_ENV['OPENAI_API_KEY'] ?? getenv('OPENAI_API_KEY');
$baseUrl = $_ENV['OPENAI_BASE_URL'] ?? getenv('OPENAI_BASE_URL');
$baseUrl = rtrim($baseUrl, '/');
if (!str_ends_with($baseUrl, '/v1')) {
    $baseUrl .= '/v1';
}

// Initialize the OpenAI client
$client = \OpenAI::factory()
    ->withApiKey($apiKey)
    ->withBaseUri($baseUrl)
    ->make();

// Store all active chat sessions
$chat_sessions = [];

// Define a common system prompt for all conversations
$system_prompt = [
    "role" => "system",
    "content" => "You are a friendly and efficient customer service attendant eager to assist customers with their inquiries and concerns."
];

// Create a new chat session with a unique identifier
function create_chat() {
    global $chat_sessions, $system_prompt;
    $chat_id = uniqid();  // Create unique session identifier
    $chat_sessions[$chat_id] = [];  // Initialize empty conversation history
    $chat_sessions[$chat_id][] = $system_prompt;  // Add system prompt to conversation history
    return $chat_id;
}

In our example, we store conversation history in an associative array called $chat_sessions, where each key is a unique chat_id. When a user sends a message, it is added to the conversation history, ensuring that the AI has access to the full context when generating a response. This approach helps create a seamless and coherent interaction between the user and the AI.