System Instructions

Introduction

Custom system instructions provide flexibility to fine-tune your PromptQL experience. This optional configuration allows you to tailor the application's behavior to your specific needs. They act as a guide for the model, shaping how it interprets questions and crafts responses.

While powerful, use this feature judiciously to maintain optimal performance.

Consider this example where we set the system instructions using the PromptQlConfig file:

kind: PromptQlConfig
version: v1
definition:
  llm:
    provider: hasura
  system_instructions: |

    You are an AI assistant that helps go-to-market teams make data-driven decisions.

    You have access to data from Salesforce (opportunities, accounts, leads), Clari (forecast categories, pipeline movement), and Postgres (custom product and revenue data). 

    Use this data to answer questions about sales performance, forecast accuracy, pipeline health, rep activity, and account trends. Prioritize clarity, brevity, and business relevance in your answers.

    If a question is ambiguous, ask for clarification.

Configuring system instructions on cloud builds

The steps above will set the system instructions for projects during local development. To set the system instructions for cloud deployments, navigate to your Settings » PromptQL and update the System Instructions field.

Best practices

When writing system instructions, your goal is to narrow the model's focus and reduce ambiguity. Here are some guidelines.

• Be clear and concise: The model performs better when instructions are direct and avoid unnecessary complexity.
• Define the role and audience: Tell the model who it is (e.g., a sales analyst, a support bot) and who it is helping.
• Specify sources of knowledge: Mention any data sources, systems, or domains the model should rely on when answering.
• Encourage appropriate behavior: Highlight priorities like brevity, clarity, formality, creativity, or error-handling.
• Provide code patterns or examples: You can include sample Python functions, formulas, or logic patterns to guide how the model generates its own code. These are not backend functions to call, but templates for the LLM to adapt and build from.
• Anticipate ambiguity: If the model might face unclear inputs, instruct it on when and how to ask for clarification.
• Set limits if needed: If certain topics, behaviors, or types of output should be avoided, explicitly mention them.
• Test and iterate: Start simple, evaluate outputs, and refine your instructions based on observed model behavior.

System instructions shape behavior

As outlined above, system instructions shape the behavior of a PromptQL application.

However, through connecting data sources and writing your own custom business logic, you're actively providing new tools which PromptQL can use. While context in the form of system instructions is helpful, you can explain with greater detail—and with much tighter scope—what each tools does using metadata descriptions.

Learn more here.