Agentic Workload for LLM: A Guide with Oxlo

Описание и применение

Создание автоматизированного агента возврата с использованием Oxlo.ai, который проверяет заказы, рассчитывает возврат и при необходимости передаёт дело человеку, обеспечивая предсказуемые расходы.

Задача

Building a refund resolution agent that verifies orders, checks policy, and decides to approve or escalates, while keeping costs predictable.

Результат

The agent correctly processes a standard refund request and escalates a hostile message, demonstrating a functional LLM‑driven workflow with predictable costs.

Рабочий процесс

• Configure Oxlo.ai client and verify connectivity
• Define mock tools and backend functions
• Write system prompt outlining workflow
• Build execution loop to handle tool calls and responses
• Add test cases for normal and hostile scenarios
• Run script and observe outputs

Доказательства

I have shipped enough internal agents to know the bottleneck is never the LLM itself. It is the loop: verify a fact, call a tool, check the result, decide what to do next.

Источник

You should see output similar to this. Exact wording varies by LLM temperature.

Источник