Building Conversational AI Models with Oxlo

Описание и применение

Создайте чат‑бота для поддержки клиентов, который использует Oxlo.ai и Llama 3.3 70B для обработки запросов о заказах и возвратах, снижая количество ручных тикетов.

Задача

Create a customer support chatbot that handles order lookups, returns, and policy questions without generating repetitive tickets for e-commerce and SaaS teams.

Результат

The chatbot successfully answers order status and return policy questions, demonstrating tool calling with Oxlo.ai’s Llama 3.3 70B model and reducing manual support tickets.

Рабочий процесс

• 1. Install Python 3.10+ and OpenAI SDK. 2. Obtain an Oxlo.ai API key. 3. Initialize the Oxlo.ai client using the OpenAI-compatible endpoint. 4. Define a concise system prompt describing the agent’s role, tone, and available tools. 5. Create a simple conversation history list that includes the system prompt. 6. Define a mock order database and a lookup_order function. 7. Specify the function schema for tool calling. 8. Build the SupportAgent class: send user messages, handle tool calls, execute lookup_order locally, and produce final natural‑language reply. 9. Test the agent with sample queries for order lookup and return policy. 10. (Optional) Deploy the agent via FastAPI or switch to a different model for multilingual/vision needs.

Доказательства

I will walk through the exact code I shipped, using Oxlo.ai's OpenAI-compatible API and Llama 3.3 70B for fast, capable reasoning.

Источник

Oxlo.ai supports function calling on Llama 3.3 70B, which makes this straightforward.

Источник