The basics

To use an LLM you need to add a “Generate text using LLMs” step to your chain. You can then choose the LLM you want to use, and set up your prompt.

The prompt input accepts both regular text and variable templating using {{}} syntax. When you start entering a variable, you will see a list of available variables to choose from. Use this functionality to enhance your prompts.

Simply use our prompt_completion transformation step. It receives a prompt param, and returns an answer.

// in your setup function
const { answer } = step('prompt_completion', {
    prompt: 'What is the best way to make a pizza?',
});

return {
    answer,
};

There are a few ways you can power up your prompts with Relevance AI: model, system_prompt and quality_control being the most powerful.